Marco Martinez

Sr Content Marketing Strategist

<p>Over the past few months, something interesting has been happening in the Coder community.</p>

<p>Developers are pushing Coder far beyond traditional workflows. They’re using it to build Minecraft mods, run personal lab-style workflows in the cloud, experiment with agent-driven systems, and move entire development setups off their laptops and into cloud workspaces.</p>

<p>I even used Coder myself to build a screenwriting app, not because it’s a traditional developer use case, but because Coder makes cloud development flexible, reproducible, and easy to move between machines.</p>

<p>What’s becoming clear is that Coder isn’t just a tool. It’s a platform that gives developers the freedom to work the way they want, while keeping environments consistent, secure, and ready to scale.</p>

<p>This flexibility matters more than ever as development moves from IDEs toward agentic workflows.</p>

<h2><strong>What we’re seeing is just the beginning</strong></h2>

<p>Most enterprises are building agentic AI on a foundation that wasn&#39;t designed for it. That&#39;s a problem.</p>

<p>AI coding tools like Cursor, Claude, and Copilot have reached mainstream adoption faster than anyone expected. What took traditional IDEs over a decade, these tools achieved in just a few years. But the workspace layer underneath them? It was built for humans, not agents.</p>

<p>That&#39;s where the community comes in.</p>

<p>At Coder, we believe the evolution from IDEs to Agentic Development Environments (ADEs) can’t be built in isolation. It has to be built with the developers, platform engineers, and contributors who are pioneering this new model. That’s why we’re focused on strengthening our relationship with the developer community in 2026.</p>

<h2><strong>Why now: The stack is changing</strong></h2>

<p>We’re seeing a new three-layer stack emerge:</p>

<ul>
<li><strong>Application layer</strong>: Chat interfaces like Cursor or Claude replacing traditional code editors<br></li>
<li><strong>Agent layer</strong>: LLMs equipped with tools to complete tasks autonomously<br></li>
<li><strong>Workspace layer</strong>: Where agents actually execute—and where Coder fits</li>
</ul>

<p>This shift is real. Teams are already pairing chat-based tools like Cursor and Claude with secure workspace infrastructure. But most workspace environments were designed for human workflows, not agent-driven ones. That means there’s a new demand for sandboxing, provisioning, and governance that isn’t being met by default.</p>

<p>This challenge can’t be solved by product alone. It needs input, experimentation, and contributions from the broader developer community.</p>

<h2><strong>How we&#39;re strengthening community</strong></h2>

<p><strong>1. The new Coder community hub</strong><br>
We launched a new <a href="https://coder.com/community">community hub</a> to make it easier to get started with Coder. Whether you&#39;re new to cloud development or installing our Community Edition on AWS, this page gives you everything you need: install Coder in 60 seconds video, setup guides, contribution instructions, and ways to engage with others building on Coder.</p>

<p><strong>2. Meetups</strong><br>
We’re hitting the road. In 2026, we’ll be hosting quarterly meetups in <a href="https://www.aicamp.ai/event/eventdetails/W2026022618">San Francisco</a>, <a href="https://www.aicamp.ai/event/eventdetails/W2026022415">Austin</a>, and <a href="https://www.aicamp.ai/event/eventdetails/W2026021910">London</a>, plus additional events in cities where the community asks us to be. These meetups are about connection, not just content. We want to hear what you&#39;re building, where you&#39;re stuck, and how we can make Coder better together.</p>

<p><strong>3. Discord + community office hours</strong><br>
Our <a href="https://discord.gg/coder">Discord</a> has been steadily growing into a place to share ideas, ask questions, and contribute. In 2026, we’re strengthening the feedback loop by bringing more of our engineers and product team directly into the conversation. We’re also launching quarterly <a href="https://discord.com/channels/747933592273027093/1387912999003754528"><strong>community office hours</strong></a>, kicking off February 12, with leaders from our product team to talk about the roadmap, recent releases, and open Q&amp;A.</p>

<p><strong>4. Champions program</strong><br>
Recognition matters. That’s why we launched the <a href="https://discord.com/channels/747933592273027093/1387912999003754528"><strong>Coder Champions Program</strong></a>. If you’re active in Discord, starring our <a href="https://github.com/coder">GitHub</a> repos, contributing templates, and helping others, we want to spotlight you. Perks include exclusive swag, access to a premium license, early access to features, and private Champions-only office hours.</p>

<p>To qualify: </p>

<ul>
<li>Participate in five community chats,<br></li>
<li>Star our GitHub repo<br></li>
<li>Help at least two people in the help channel.<br></li>
<li>Nominate yourself or a fellow community member in our Discord <a href="https://discord.com/channels/747933592273027093/971231372373033030">general</a> channel.</li>
</ul>

<p><strong>5. Community-led contributions</strong><br>
Whether you&#39;re submitting PRs, sharing modules, or improving templates, we want Coder to be an open canvas for your ideas. We&#39;re working on lowering the barrier to entry for contributors, including new guides for first-time PRs and easier ways to get your contributions noticed.</p>

<p><strong>6. Devolution podcast</strong><br>
Hosted by our Field CTO, Nicky Pike, <a href="https://www.youtube.com/@Devolution-Podcast">The <strong>[Dev]olution</strong> podcast</a> is where we highlight voices from the industry and our community, building real-world systems with Coder. From <a href="https://youtu.be/cZ8mElnep08">Minecraft mods</a> to agent-powered tooling, we’re featuring the creators who are pushing the boundaries of what’s possible.</p>

<h2><strong>We can’t build this alone</strong></h2>

<p>The future of development is agentic. But the infrastructure to support it isn’t just technical, it&#39;s social. We need awareness, education, and collaboration across the ecosystem. </p>

<p>Community is how this work gets done.</p>

<p>We’re betting that the best ideas won’t come from inside our company, they’ll come from you.</p>

<h2><strong>How to get involved</strong></h2>

<p>If you&#39;re using Cursor, Claude, or any of the new AI dev tools and hitting governance walls, we want to hear from you. If you&#39;re building agent workflows and need secure infrastructure, check out our open source tools. If you&#39;re just curious what others are doing, join our <a href="http://coder.com/chat">Discord</a>.</p>

<p>And if you want to help build the future of agentic development?</p>

<p>We&#39;re hiring. See open roles at coder.com/careers.</p>


Our community is pushing Coder far beyond traditional workflows. They’re using it to build Minecraft mods, run personal lab-style workflows in the cloud, experiment with agent-driven systems, and move entire development setups off their laptops and into cloud workspaces.

Building Together: Coder's Community Vision for 2026

<p>The first message someone gets in a community can make or break whether they stick around. At Coder, we’re a community-centric company, which means those early interactions are not just nice to have. They are foundational to how we operate. When someone joins our <a href="http://coder.com/chat">Discord</a> and introduces themselves in #intros, that moment matters. It’s where they decide if this space actually values them or if they’re just another username scrolling by.</p>

<p>As the community grew, that moment started to slip.</p>

<p>The #intros channel was moving faster than I could keep up with. New people were joining every day, sharing who they are, what they’re building, and why they’re here. I wanted to give everyone a thoughtful welcome, but my response time kept slipping due to travel, shifting event priorities, last-minute projects, and the general chaos of startup life. Some people got a quick “hey, welcome in, what are you building?” if I happened to be online. Others waited hours. None of that reflected how much we actually value the people who choose to spend time in our community.</p>

<p>At the same time, I didn’t want a bot firing off generic replies on my behalf. I still wanted to be part of the welcome. I just didn’t want to be glued to one channel all day.</p>

<p>So I kept coming back to the same idea.</p>

<p>What if something else watched the channel for me and drafted the replies?</p>

<p>That was the moment I realized I didn’t just need automation. I needed an agent I could trust, control, and stay involved with.</p>

<p>I could still be the human touch at the end, but I wouldn’t be tied to Discord. I wanted an agent watching that channel around the clock, drafting responses based on what people actually wrote, then sending them to me in Slack so I could approve or edit before anything was posted.</p>

<p>That idea turned into a real project. I’d built a small to-do app with Blink before, so I knew this wasn’t impossible. The difference this time was that the stakes were real. This wasn’t a weekend experiment. This was me trying to solve a daily problem that directly affected whether new members felt welcomed or overlooked.</p>

<h2><strong>How I built the first version of my AI Discord agent using Blink</strong></h2>

<p>I started the same way I did last time. I opened <a href="http://blink.so">Blink</a>, an AI coding agent built by Coder that lets you build real software by describing what you want in plain language, and said something like, “Help me build a Discord bot that welcomes new members.”</p>

<p>Blink walked me through it step by step. File by file. Package by package. Slack setup. Discord setup. GitHub repo creation. It felt less like coding and more like having a developer sitting next to me saying, “type this, here’s why, now try this.”</p>

<p>Before long, I had the whole thing running on my laptop.</p>

<p><img src="https://www.datocms-assets.com/19109/1766433185-ai-community-blog_1.png" alt=""></p>

<p>The workflow was smooth. Someone posts an intro in Discord, I get a Slack message with a suggested reply, I tap approve, edit, or skip, and it posts to Discord as a thread. Seeing it actually work felt pretty incredible.</p>

<p>But there was a catch.</p>

<p>If I closed my laptop, the bot stopped.</p>

<p>If I lost WiFi, the bot stopped.</p>

<p>If I walked away from my desk, the bot stopped.</p>

<p>So technically it worked, but it didn’t solve the underlying consistency problem at all.</p>

<h2><strong>Using Ollama to generate AI responses locally for the Discord bot</strong></h2>

<p>One early surprise was the AI itself. I assumed I’d need an external API for responses, but Blink suggested running a model locally using Ollama. <a href="https://ollama.com/">Ollama</a> is a local AI runtime that lets you run language models on your own machine instead of sending data to an external service.</p>

<p>I installed it, pulled the phi3.5 model, and suddenly the bot was generating responses with no API keys and no data leaving my environment. I looked at a few different options along the way, Llama, Phi, Mistral, but landed on phi3.5 after reading a mix of articles, Google searches, and community feedback that pointed to it being solid for conversational tasks with a reasonable model size.</p>

<p>That choice ended up being exactly what I hoped for. Fast local AI without calling a third-party server. It ran quickly, it was private, and it made it very clear how far open source models have come for real work.</p>

<p>Small models have quickly caught up in performance and usability. According to research from <a href="https://hai.stanford.edu/ai-index/2025-ai-index-report/technical-performance">The Stanford Institute for Human-Centered AI (HAI)</a>, Microsoft’s Phi-3-mini, with just 3.8 billion parameters, reached the same level of performance with another local, small model that was 142 times its size. That kind of progress is what makes fast, private, on-device AI a practical solution, not just theoretically possible.</p>

<p><img src="https://www.datocms-assets.com/19109/1766433185-ai-community-blog_2.png" alt=""></p>

<p>But the same problem came back.</p>

<p>If the laptop sleeps, everything sleeps.</p>

<p>I needed somewhere the agent could live permanently.</p>

<h2><strong>Deploying the AI Discord agent to a Coder Cloud Workspace (and what broke along the way)</strong></h2>

<p>I’d talked about Coder for years at events and meetups, but this was the first time I actually needed it like a normal user. I wasn’t looking for something to magically host my app for me. I needed a real cloud environment where my now production-ready agent could live outside my laptop while I figured out what it actually takes to keep something running continuously.</p>

<p><a href="https://coder.com/cde/premium">Coder</a> gives you cloud development workspaces that stay accessible whether your laptop is open, closed, or nowhere nearby. That made it a natural next step for moving this agent out of my local environment and into the cloud, without pretending I’d suddenly solved production infrastructure.</p>

<p>I logged into Coder, spun up a workspace, installed Node and Ollama, pulled the repo, configured the Discord and Slack credentials, and started the agent with Blink guiding me through the process. </p>

<p><img src="https://www.datocms-assets.com/19109/1766433185-ai-community-blog_3.png" alt=""></p>

<p>Everything connected immediately. Discord. Slack. Ollama. Logs streaming. And most importantly, the agent kept running even when I closed my laptop and walked away, at least long enough to prove this could live outside my local machine.</p>

<p>Coder didn’t make the agent production-hosted for me, but it gave me a persistent cloud environment where I could start learning what “always on” actually means.</p>

<p>That’s when it clicked.</p>

<p>Building something locally is one thing.</p>

<p>Keeping it running continuously is a completely different skill.</p>

<p>The development part was fairly straightforward. The operational part got messy fast.</p>

<p>A few days later, Slack notifications were still coming in, but the AI suggestions were gone. Ollama was throwing connection errors. The workspace had restarted, and Ollama never came back up. I’m still not totally sure why the restart happened in the first place. This wasn’t a Coder issue. It was my lack of understanding how to structure the environment and how long I actually wanted it to stay alive.</p>

<p><img src="https://www.datocms-assets.com/19109/1766433185-ai-community-blog_4.png" alt=""></p>

<p>That single hiccup taught me more about infrastructure than the entire initial build. I learned how to check background processes, inspect logs, restart services, and why startup sequencing matters. I also learned that if you try to automate a startup script before everything is installed, you can brick your entire workspace. I did exactly that. Blink and I had to rebuild the whole thing. Vibe coding will humble you.</p>

<p>Looking back, I probably should have used a proper system service or something similar. But learning through breakage forced me to understand what was actually happening under the hood. There’s something about watching your own environment fail that makes you actually read the documentation instead of skimming it.</p>

<h2><strong>Key lessons from running an AI-powered Discord agent in production</strong></h2>

<p>Once everything settled and had been running for a few weeks, I started noticing patterns that went beyond my Discord workflow.</p>

<p>The agent doesn’t replace me. I still review everything. But I’m not jumping into Discord every twenty minutes anymore. It handles the monitoring and drafting, and I handle the approval and personal touch. The human part stays human, which was the whole point.</p>

<p>The AI itself was the easy part. Keeping everything running reliably was where I spent most of my time. The model generates responses just fine. The real challenges are making sure Ollama starts correctly, Slack and Discord connections stay alive, and the system recovers when something inevitably fails. The bottleneck isn’t intelligence. It’s operations.</p>

<p>Running phi3.5 locally also means no API bills, which is nice, but the bigger thing is that intro messages aren’t leaving my infrastructure. I didn’t think I’d care about that much at first, but now it feels important. Privacy and cost control matter a lot more once you actually ship something real.</p>

<p>The wildest part is that I built this at all. I don’t have a CS degree or formal training. Blink wrote most of the code. Coder gave me somewhere reliable to run it. A year ago this would have felt completely out of reach. Now people without traditional dev backgrounds can build tools that genuinely help teams without waiting months for engineering resources.</p>

<p>And this pattern clearly extends beyond community work. Support teams, ops teams, developer relations, anyone dealing with high-volume communication could use something like this. The specifics would change, but the workflow wouldn’t. You’re not automating the decision. You’re automating the grunt work around the decision so humans can focus on what actually requires judgment.</p>

<h2><strong>Why AI agents will reshape community management and developer workflows</strong></h2>

<p>Blink made it possible to build the bot. Coder made it possible to run and iterate on it in the cloud. Ollama kept everything private and cost-effective. Together they created a pattern where people with ideas can solve real problems without getting stuck in a backlog somewhere.</p>

<p>I think that’s where AI is quietly heading. Not massive automation projects or polished demos that break in production. Real agents solving real, repetitive problems that make everyday work easier and more consistent.</p>

<p>That shift is already showing up in how teams are actually using AI. According to <a href="https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-state-of-ai">a recent McKinsey survey</a>, 88% of organizations report using AI in at least one business function, but most of those deployments are focused on narrow, task-specific use cases rather than large-scale automation. The value is coming from practical agents that improve everyday workflows, not from replacing entire functions.</p>

<p><img src="https://www.datocms-assets.com/19109/1766433185-ai-community-blog_5.png" alt=""></p>

<p>If you work with communities, you already know consistency and connection matter. AI can’t replace that, but it can support it. It gives you time back. It filters noise. It helps you scale the moments that actually make people feel welcome.</p>

<h2><strong>Final thoughts on building a 24/7 AI agent as a community manager</strong></h2>

<p>Looking back, the whole project took about a week of real work, most of it spent learning how to keep things running reliably. What surprised me wasn’t how hard the AI part was, but how much there is to learn once something needs to live outside your laptop.</p>

<p>If you’re dealing with similar scaling problems, I hope this helps. And if you end up building something similar, let me know what you learn in <a href="http://coder.com/chat">our Discord community</a>. I’m always curious what other people automate when no one is watching.</p>

<p>If you want the full backstory on how I got started with this approach, I wrote about my first vibe coding project here:<br>
<a href="https://coder.com/blog/vibe-coding-my-first-app-with-blink-no-cs-degree-needed">**Vibe Coding My First App with Blink (No CS Degree Needed)</a>**</p>


When my Discord community started growing faster than I could keep up with welcome messages, I needed a solution that wouldn't sacrifice the personal touch. This is the story of building a 24/7 AI agent that monitors our intros channel, drafts personalized responses, and sends them to me in Slack for approval before posting. Along the way, I learned that building the AI part was straightforward. Keeping it running reliably in the cloud was where the real education happened. If you're dealing with scaling problems in community management, support, or any high-volume communication work, this breakdown shows what's actually possible when you combine local AI models, cloud development environments, and a willingness to learn through breakage.

How I Used AI to Solve a Real Community Scaling Problem

<p>I am not a software engineer. I have never taken a formal computer science course. I cannot write Python from scratch, and I only recently started exploring Linux. But after two years at <a href="http://www.coder.com">Coder</a>, I have learned how cloud development environments work, how they support modern teams, and how secure on demand workspaces remove friction from the development process.</p>

<p>Despite my lack of formal training, I was able to prototype the first version of a real working to do app that fits my own workflow and preferences by working side by side with an AI agent called Blink.</p>

<p>This was not a drag and drop project. It was real code, version controlled in GitHub, deployed through Vercel, and powered by a live backend in Supabase. Everything ran inside a secure Coder workspace.That foundation gave me the confidence to try something new. </p>

<p>This is the beginning of my vibe coding journey. It is my attempt to learn by building and to work with AI as a partner instead of writing every line myself.</p>


<h1><strong>What is vibe coding? And, what does it mean to me?</strong></h1>

<p>Vibe coding is a way of building software by describing what you want to create instead of writing every line by hand. I can say something straightforward like “Create a to-do list with categories and deadlines,” or something more abstract like “Make this feel more like the Things Three app from Culture Code.” I can focus on ideas, flow, and experience while the agent handles the structure and syntax.</p>

<p>Vibe coding is not about skipping fundamentals. It removes the early blockers that usually stop someone like me from getting started. The more I build this way, the more curious I become about what is happening behind the scenes. Working with Blink, an open source AI agent building tool from Coder, has pushed me to learn Git, CSS, and even take Linux classes because I want to understand how everything connects under the hood.</p>

<p>For me, vibe coding is a new way in. It gives me a place to start and a safe way to learn through real projects, not theory.</p>


<h1><strong>What’s different about Blink?</strong></h1>

<p>Blink is an AI agent that runs inside a secure cloud development environment powered by Coder. It has access to my GitHub repo, my project history, and the tools needed to deploy real applications. It understands context and builds software in response to my prompts.</p>

<p>Before Blink, I experimented with Coder Tasks and Claude Code. Those early experiments gave me an exciting preview of how cloud infrastructure and AI can work together. I was able to generate simple prototypes and understand how version controlled AI tasks could shape early ideas.</p>

<p>Blink pushed that experience much further. When I created my workspace, GitHub, Vercel, and Supabase were already connected and ready. I did not write config files or perform setup steps. I just described what I wanted, and Blink guided me from idea to deployment.</p>

<p>Sometimes things broke. There was a point where menus shifted when I implemented a code change or when I tried to update subtasks to respond a certain way but the tasks broke. I ran into these kinds of bugs during development and at one point I asked Blink to wipe the entire build and start from scratch while keeping my GitHub repo, Supabase project, and Vercel instance intact. It did exactly that. Clean slate. Same infrastructure. I still had bugs but I was able to fix them with Blink. </p>

<p>Blink is not a one off code generator. It is a persistent partner that stays with you as the project grows.</p>


<h1><strong>How does Blink compare to other AI coding tools?</strong></h1>

<p>Blink is not the only agent you can use inside Coder. My first experience was with Claude Code, a general purpose coding agent developed by Anthropic. Claude is powerful and flexible, and I used it through Coder Tasks to run scoped technical work inside secure cloud environments. That setup helped me get early prototypes running and showed me how agents can shape real development workflows when they are backed by the right infrastructure.</p>

<p>Coder Tasks is the layer that makes this possible. It lets you run agents like Claude inside isolated cloud workspaces that are version controlled and fully contained within your own environment. You can give an agent a specific prompt, scope it to a branch, and see exactly what it changes.</p>

<p>Blink works differently. It is built to go deeper into the development lifecycle. It remembers context, understands the state of your project, and can stay with you as you iterate. Claude Code excels at one off tasks and suggestions. Blink feels closer to a project partner. It can track progress, make decisions, and help you move from idea to deployed app in a more continuous way.</p>

<p>So it is not that Blink is more powerful than Claude Code or other agents. It is more specialized. It is built for people like me who want to create real applications through conversation inside a secure, consistent development environment.</p>


<h1><strong>Who is vibe coding for?</strong></h1>

<p>One of the unexpected highlights of this journey was connecting with someone else who was also vibe coding with Blink. Natalie Lambert, the founder and managing partner of <a href="https://genedge.co/">GenEdge</a>, built something completely different – a lightweight CRM for her startup. But even though our projects were different, our experiences had so many commonalities.</p>

<p>We both had the same realization: vibe coding isn&#39;t just for people who can&#39;t code. It&#39;s for people who have ideas and want to see them work without waiting for a developer to have bandwidth.</p>

<p>That doesn&#39;t eliminate the need for skilled engineers. It just gives them a better starting point. When I hand off something I vibe coded, it&#39;s already past the sketch phase. It has structure, it runs, and it reflects what I&#39;m actually trying to build.  Blink makes that kind of collaboration feel possible, even for people who are not traditional engineers.</p>


<h1><strong>What I Learned Through Vibe Coding</strong></h1>

<p>Every iteration taught me something new. I learned how small UI changes can affect an entire layout, how frontend and backend work together, and why scoped increments produce better results than sweeping requests. I learned to commit versions carefully, test in branches, and take smaller steps.</p>

<p>The biggest shift happened when I started asking Blink for guidance rather than directives. Instead of saying “Build this,” I would ask “What would you recommend here?” That change created a more collaborative process. Blink showed me trade offs and approaches I would not have considered.</p>

<p>This style of building made software concepts feel less intimidating and more approachable. It taught me terminology, best practices, and version discipline through direct experience. The learning did not come from documentation. It came from doing the work with Blink in the loop.</p>


<h1><strong>What’s next for my vibe coding journey? How I work with Blink today</strong></h1>

<p>The to-do app I built is still evolving. I&#39;m refining the UI, testing new features, and learning better habits every time I open the workspace.</p>

<p>But the bigger shift is this: vibe coding has made me more confident, more curious, and more eager to keep learning. It hasn&#39;t flattened the learning curve. It&#39;s just made it more accessible.</p>

<p>That&#39;s what tools like Blink enable. Not shortcuts around the fundamentals, but a way in for people who wouldn&#39;t have started otherwise.</p>

<p>And honestly? That feels like the point.</p>

<p>If you want to try vibe coding yourself, you can sign up for early access to <a href="https://blink.so/">Blink</a>.</p>


I am not a software engineer. I have never taken a formal computer science course. I cannot write Python from scratch, and I only recently started exploring Linux. Despite my lack of formal training, I was able to prototype the first version of a real working to do app that fits my own workflow and preferences by working side by side with an AI agent called Blink.

Vibe Coding My First App with Blink

Vibe Coding My First App with Blink (No CS Degree Needed)

<p>Are you willing to be slower today to be faster next year?</p>

<p>In the most recent episode of <a href="https://www.coder.com">Coder</a>’s <a href="https://www.youtube.com/watch?v=br8ymXaHeKE">[DEV]olution podcast, Are We Shipping Code We Don’t Trust? The AI Productivity Paradox</a>, Jason Baum, Project Leader at Selenium, discussed what happens when AI promises speed but delivers friction instead. The conversation raised a critical question that platform engineering teams everywhere are wrestling with: why are experienced developers slowing down with AI tools, and what does that mean for organizations investing in these technologies?</p>

<p>At Coder, we see this pattern daily. Teams adopt AI coding agents expecting immediate velocity gains, only to watch productivity dip in the first few months. The culprit is not the AI itself; it&#39;s the infrastructure beneath it. When your development environments can&#39;t keep pace with AI experimentation, you can&#39;t tell whether the slowdown is the learning curve or your systems breaking under new patterns of use.</p>


<h1>The paradox of &quot;faster&quot; tools</h1>

<p>Every developer loves a shortcut. So when AI coding tools showed up promising to write less, ship faster, and automate the boring stuff, of course developers leapt at the chance to implement them into their work.</p>

<p>Then the data came in, and it was weird. Studies showed that experienced developers were actually slower when using AI. <a href="https://metr.org/blog/2025-07-10-early-2025-ai-experienced-os-dev-study/">METR</a> found a 19-percent decrease in speed among senior engineers working on familiar codebases. People got confused. Some got defensive. How could tools built to speed things up end up slowing them down?</p>

<p>Here&#39;s the thing though: this happens every time we change how we build software. Version control, automated testing, CI/CD pipelines. All of them made us slower at first. Teams had to pause, learn, deal with the friction. The AI slowdown is just the latest version of that. Developers aren&#39;t getting worse; they&#39;re relearning how to think. And that always takes longer than anyone wants to admit.</p>


Source: METR, “Against Expert Forecasts and Developer Self-Reports, Early-2025 AI Slows Down Experienced Open-Source Developers.”

<h1>Why the slowdown happens</h1>

<p>AI changes the rhythm of how you actually work. You used to be both creator and reviewer. You wrote the code, you caught your own mistakes. Now you&#39;re more like a collaborator. You prompt the model, review what it gives you, test it, then decide: keep it, tweak it, or throw it out.</p>

<p>The code shows up instantly, but figuring out if it&#39;s actually good is what takes time and where things get messy. Excited developers that were quick to use generative AI coding assistants often found themselves falling flat as they got bogged down in low-quality code or code that seemed fine but ultimately failed in production. </p>

<p>Jason Baum illustrated this on our [DEV]olution podcast in a way that stuck with me: &quot;We&#39;re running before we walk with AI.&quot; </p>

<p>Developers are still figuring out the pacing. When is the model right? When is it confidently wrong? When has it just completely lost the plot?</p>

<p>Last quarter, I watched a team go through this exact phase. Their first few sprints with AI-assisted tooling felt clumsy. They spent more time debugging than building. But by the third sprint, something clicked. Their reviews got tighter. They started spotting issues faster. The slowdown wasn&#39;t failure; it was just what learning looks like when you can actually see it happening.</p>


<h1>The trust gap</h1>

<p>Trust sits at the center of all this. <a href="https://survey.stackoverflow.co/2025/ai/?utm_source=chatgpt.com">Stack Overflow&#39;s 2025 Developer Survey</a> found that 84 percent of devs already use AI tools, but only 33 percent trust their results. That&#39;s a massive gap.</p>

<p>And it&#39;s not fear of the tech. It&#39;s that nobody really knows how it&#39;s going to behave. Devs trust things they understand. They trust compilers because they are deterministic and predictable. They don&#39;t trust probabilistic models that might give you perfect code one minute and introduce a silent bug the next. And honestly? That skepticism is healthy.</p>

<p>Closing that gap takes repetition, not marketing. Developers need to see patterns of reliability over time. I&#39;ve seen teams build that trust slowly by running AI-generated code through the same review pipelines they use for everything else. At first, every suggestion got torn apart but soon they were merging small changes without anxiety. The process built trust because it looked like every other engineering habit: test, review, verify, repeat.</p>

<p>The goal isn&#39;t blind faith. It&#39;s confidence through verification. Unit tests built trust in refactoring the same way, by making the invisible visible.</p>


Source: Stack Overflow, 2025 Developer Survey — AI tools in the development process.

<h1>The long view</h1>

<p>Six months from now, this curve starts to flatten. Devs who stick with AI tools will move past the awkward discovery phase into something closer to fluency. They&#39;ll learn to frame prompts like they design APIs: clear inputs, defined expectations, repeatable outcomes. They&#39;ll figure out which parts of the stack actually benefit from AI and which parts still need a human making the call.</p>

<p>Jason put it well on the podcast: &quot;The best developers aren&#39;t the ones who avoid new tools. They&#39;re the ones who learn how to question them.&quot; It’s a simple line, but it captures the whole point of this phase: learning how to think again before we automate everything.</p>

<p>That mindset is what separates teams that really integrate AI from teams that just play with it. The advantages show up quietly at first: documentation gets better, feedback loops tighten, test coverage improves. But it compounds. Developers who spent hours debugging generated code will shift to designing systems faster with fewer regressions.
The ones willing to be &quot;slow&quot; now will be the fastest later.</p>


<h1>Redefining &quot;productivity&quot;</h1>

<p>We&#39;ve measured productivity in software the same way for decades: lines of code, commits, tickets closed. AI breaks that completely. When a model can spit out hundreds of lines in seconds, output stops meaning anything. What actually matters is how fast you can go from idea to a reliable, maintainable solution that needs understanding, not volume.</p>

<p>The slowdown happening right now is that transition. Devs are putting in time upfront to use AI responsibly so they can move faster later without breaking things. The metric that&#39;ll matter going forward is probably something like cognitive velocity: how fast teams can understand problems, evaluate solutions, and deliver stable results.</p>

<p>It&#39;s slow at first because comprehension takes time. But once that becomes muscle memory, the multiplier effect is going to be huge.</p>


<h1>Why infrastructure is an AI enabler</h1>

<p>At Coder, this is the shift we&#39;re built for. Organizations adopting AI need infrastructure that supports exploration without risking production. Self-hosted cloud development environments give developers and AI agents the same consistent, secure workspace every time. </p>

<p>Jason called this out on the podcast: &quot;When companies start giving AI agents real access to environments they don&#39;t fully control, they&#39;re opening the door to chaos. It&#39;s not just a productivity issue. It&#39;s a security one.&quot;</p>

<p>He&#39;s right. The risk isn&#39;t just that AI slows you down, it&#39;s that inconsistent infrastructure makes it impossible to diagnose issues, trust results, or maintain security posture. When AI agents start autonomously generating and executing code, you need environments you can govern completely.</p>

<p>This is why consistent, production-like workspaces matter more than ever. They provide a predictable foundation that fosters faster, safer learning. The goal is twofold:  speed and stability. You need to build trust in new workflows while maintaining control over what&#39;s running where, who has access, and what data is exposed.</p>

<p>Coder provides exactly this: reproducible <a href="https://coder.com/blog/ephemeral-environments">ephemeral environments</a> that developers and AI agents can spin up in seconds, all running on infrastructure you control. Whether you&#39;re in financial services managing compliance requirements or a tech company scaling AI adoption across hundreds of developers, you get the security of self-hosted infrastructure with the speed of cloud-native tooling. The result is every workspace used by both human developers or AI agents starts from the same known-good state. No configuration drift. No &quot;works on my machine&quot; surprises. That consistency is what makes AI experimentation possible without the chaos.</p>


<h1>The real lesson</h1>

<p>Every big shift in software development starts with discomfort. The tools promise speed, but using them the first time feels like walking through mud. The devs who push through, who are willing to be patient while they figure this out, end up faster, better, more confident on the other side.</p>

<p>AI isn&#39;t killing developer productivity. It&#39;s redefining what productivity even means. The slowdown isn&#39;t a problem; it&#39;s the investment you make to actually get somewhere.</p>

<p>What surprised me most in all these conversations? It&#39;s not that AI makes developers slower. It&#39;s that the best teams are choosing to be slower right now. They&#39;re deliberately taking time to build the right habits, even when it feels inefficient. That kind of patience is rare in an industry that treats velocity like a religion.</p>

<p>So, I’ll ask again: Are you willing to be slower today to be faster next year?</p>

<p>Because that&#39;s the actual trade-off, and I&#39;m not convinced everyone realizes they&#39;re making it.</p>


<h1>Listen to the full conversation</h1>

<p>Want to hear more about why AI tools create friction before they deliver value and what this shift means for software development? Tune in to the[DEV]olution podcast episode with Selenium’s Jason Baum, who knows more about developer productivity and testing than most people forget, to dig into all of this.</p>

<p>You can find the full episode on <a href="https://youtu.be/br8ymXaHeKE?si=gO9b6AiPTBoLnrN8">YouTube</a>, <a href="https://open.spotify.com/episode/6usd12PVSVY8SI0BCeXnNg?si=ufokcWjNSVKB869FHyAMyA">Spotify</a>, or <a href="https://podcasts.apple.com/us/podcast/are-we-shipping-code-we-dont-trust-the-ai/id1839475248?i=1000730782333">Apple Podcasts</a>.</p>


Experienced developers often slow down when adopting AI coding tools, not because the technology fails, but because they're learning to shift from solo creator to collaborative reviewer of AI-generated code. Drawing from data showing a 19% velocity decrease among senior engineers and insights from the [DEV]olution podcast, this piece explores why the initial friction is actually an investment. Teams willing to be slower today while building trust through verification and establishing the right infrastructure will ultimately be faster, more confident, and more productive tomorrow.

The AI Learning Curve: Why Devs Get Slower Before They Get Faster

<p>&quot;But it works on my machine.&quot;</p>

<p>Every developer has said this. Most of us have said it more times than we&#39;d like to admit. When it&#39;s just humans writing code, that gap between local and production is annoying but manageable. You debug it, you fix it, you move on.</p>

<p>But now we&#39;re handing that same messy local environment to AI agents: tools that can write files, execute commands, and make decisions without asking permission first. That gap isn&#39;t just annoying anymore. It&#39;s actually dangerous.</p>

<p>Ben Potter, our VP of Product here at <a href="https://www.coder.com">Coder</a>, put it perfectly on our first episode of the <a href="https://www.youtube.com/watch?v=1I9OS3bjNRE"><em>Devolution</em> podcast: Your local machine might be lying</a>. As Ben puts it, &quot;the code we write there is not what actually runs in production, and the gap between the two only widens when AI agents enter the mix.&quot;</p>

<p>The solution isn&#39;t better prompts or smarter models. It&#39;s better infrastructure. Short-lived environments that give both humans and AI agents a consistent, production-like place to work.</p>


<h2>What we mean by AI agents</h2>

<p>Quick clarification: when I say &quot;AI agent,&quot; I&#39;m not talking about glorified autocomplete. I mean a language model that&#39;s connected to actual tools and can act on feedback. It writes files, runs terminal commands, reads the output, and adjusts what it does next. These things are semi-autonomous. They don&#39;t wait for you to approve every action.</p>

<p>Which is great for productivity. And also means they can cause a lot of damage if they&#39;re working in a broken or unpredictable environment.</p>


<h2>Your laptop is lying to you</h2>

<p>Local dev environments are all different. Every single one. You&#39;ve got different OS versions, different installed libraries, config files you forgot about three years ago, dependencies that got pinned to weird versions to work around some bug that&#39;s probably been fixed by now.</p>

<p>Devs deal with this by working around it. They disable security checks that are &quot;too annoying.&quot; They hardcode paths. They also accumulate technical debt like it&#39;s a hobby.</p>

<p>For humans, this is frustrating but survivable. For AI agents, it&#39;s completely unsustainable. These tools need stability and reproducibility. Asking an agent to operate in your janky local setup is like asking someone to learn to drive in a car where the steering wheel only sort of works sometimes.</p>


<h3>Ephemeral environments fix this</h3>

<p>Ephemeral environments are pretty elegant. Every time you spin one up, it&#39;s completely fresh with no leftover files, no mystery configuration from last week&#39;s debugging session. They mirror production closely, so what you&#39;re testing actually reflects what will run in the real world.</p>

<p>And when something goes wrong, you don&#39;t spend hours trying to clean up the mess. You just delete the environment and start fresh.</p>

<p>At Coder, we&#39;ve built our platform around ephemerality. Every dev, every AI agent gets launched into a clean workspace that behaves like production. No more &quot;well, it works on my machine&quot; because everyone&#39;s machine is essentially the same.</p>

<p>Here&#39;s what happened to me last week: I was testing an agent that got stuck in a loop and started generating a bunch of garbage code. In a traditional local environment, I&#39;d be stuck cleaning that up manually, trying to figure out what got changed, maybe even dealing with a corrupted state. Instead, I hit delete, spun up a new workspace, and was back to work in maybe 30 seconds.</p>


<h2>Building guardrails that actually work</h2>

<p>AI agents don&#39;t second-guess themselves. If they decide to run rm -rf, they&#39;re just going to do it. You need real guardrails, not just good intentions.</p>

<p>Short-lived workspaces are a big part of that safety net. They contain the blast radius when something goes wrong. Combine that with sensible defaults (read-only filesystems where appropriate, no internet access unless explicitly needed), and you&#39;ve got a workspace where agents can operate without the risk of catastrophic mistakes.</p>

<p>This isn&#39;t just us saying this. Anthropic&#39;s documentation explicitly recommends strict permissions for AI agents. Simon Willison has written about what he calls the &quot;<a href="https://simonwillison.net/2025/Jun/16/the-lethal-trifecta/">lethal trifecta</a>&quot;: AI agents with too much access in poorly secured environments. Ephemerality cuts through that risk.</p>


<h2>The human benefits</h2>

<p>This isn&#39;t just about making AI safer. Human developers benefit from this too.</p>

<p>Clean-slate environments make it easier to handle sensitive data, keep costs under control, and stay compliant with various regulations. But honestly, the biggest win is just eliminating the &quot;works on my machine&quot; problem entirely.</p>

<p>When everyone&#39;s working from the same production-like environment, debugging becomes about the actual code instead of someone&#39;s weird local setup. New developers can onboard faster because there&#39;s no multi-day setup process. Security teams finally get the audit trails and consistency they&#39;ve been asking for.</p>


<h2>Making this work at scale</h2>

<p>At a small scale, ephemeral environments are a nice quality-of-life improvement. At enterprise scale with thousands of devs, dozens of AI agents all running simultaneously, they become essential infrastructure.</p>

<p>But you can&#39;t just wing it. You need proper orchestration to handle provisioning and teardown. You need cost controls so you don&#39;t wake up to a surprise AWS bill.</p>

<p>And you need observability to understand what your agents are doing and when they did it.</p>

<p>The hard part of deploying AI at scale isn&#39;t the AI itself. It&#39;s doing it in a way that&#39;s safe, predictable, and auditable across an entire organization.</p>


<h2>Where we go from here</h2>

<p>If you&#39;re investing in AI tooling but haven&#39;t thought about your infrastructure, you&#39;re probably taking on more risk than you realize. Ephemeral environments aren&#39;t optional. They&#39;re foundational.</p>

<p>AI agents are getting better fast, but they&#39;re not magic. They need good infrastructure to be reliable. With ephemeral workspaces, you can move faster, stay safer, and actually scale this stuff with confidence.</p>

<p>What surprised me most about working on this infrastructure problem is how much it benefits everyone, not just the AI. We built Coder to solve the agent problem, but our customers keep telling us the human developers are just as happy to finally be free of local environment hell.</p>

<p>If you want to see what this looks like in practice, check out Coder. We&#39;ve built our platform around giving developers and AI agents the guardrails they need without slowing them down.</p>

<p>Want to hear the full conversation with our VP of Product, Ben Potter? You can catch the full episode on <a href="https://youtu.be/1I9OS3bjNRE">YouTube</a>, <a href="https://open.spotify.com/episode/2Ypeffrbr8pVbQhsbb4qTC?si=Ungx-7vySQm809iA2TghSA">Spotify</a> or <a href="https://podcasts.apple.com/us/podcast/your-local-machine-might-be-lying-ben-potter-on-ai/id1839475248?i=1000728194573">Apple Podcast</a>.</p>


Local development environments are inconsistent and unpredictable, filled with forgotten configs and custom workarounds. For human developers, this is frustrating—but for AI agents that can autonomously execute commands, it's genuinely dangerous. Ephemeral environments solve this by providing fresh, production-like workspaces that can be spun up instantly and deleted just as quickly, containing the blast radius when things go wrong. Beyond making AI agents safer, they eliminate the "works on my machine" problem entirely, helping human developers debug actual code instead of investigating mysterious local setups.

Why your AI agents need ephemeral environments (and your devs probably do too)

Why Your AI Agents Need Ephemeral Environments (And Your Devs Probably Do Too)

<p>We recently hit an incredible milestone: <strong>100,000 GitHub stars across all Coder projects.</strong> This moment is more than just a number. It’s a celebration of the developers, contributors, and community members who’ve helped shape the future of cloud-based development.</p>

<p>From day one, our mission at Coder has been clear: <strong>empower developers with secure, consistent, and flexible cloud development environments.</strong> Reaching this milestone reinforces that the open source community believes in that vision too.</p>


<h2>Open source at the heart of Coder</h2>

<p>We launched the <a href="https://coder.com/cde">Coder Community Edition</a> to give developers full access to Coder on their own terms, with no license keys, no trial subscriptions, and no sales hoops. From the start, it was open source and fully self hosted, designed to let developers deploy, inspect, and customize every part of the platform.</p>

<p>Coder’s core codebase is written in Go, and every CLI, UI, and backend component is open and extensible. Developers can audit the code, report bugs, suggest features, and contribute improvements in the open.</p>

<p>Since then, thousands of developers from indie builders to large teams have adopted Coder and become part of the movement. We now maintain more than 100 public repositories, from prebuilt templates to automation tools. At the center is <a href="https://github.com/coder/code-server">code-server</a>, which brings VS Code to the browser and powers development environments around the world.</p>

<p>And our community is not just along for the ride. They are actively shaping the platform. Contributors like <a href="https://github.com/benz0li">Olivier Benz (benz0li)</a> have played a key role in maintaining code-server’s compatibility with upstream VS Code releases, pushing dozens of updates to keep it current. Others, like <a href="https://github.com/jawnsy">Jonathan Yu (jawnsy)</a>, have improved CI workflows, build tooling, and version handling, while <a href="https://github.com/SuperSandro2000">Sandro Jaeskel (SuperSandro2000)</a> has helped optimize Docker images and streamline container usage.</p>

<p>These are just a few of the many developers helping drive Coder forward in ways both visible and behind the scenes, shaping a platform that belongs to all of us.
Open source is not just how we build. It is who we are.</p>


<h2>Join the Coder community</h2>

<p>Coder is powered by people and there is a place here for every kind of contributor. Whether you are a seasoned engineer, an open source enthusiast, or someone just getting started, your input matters. From sharing feedback and feature ideas to building templates, helping others troubleshoot, or simply asking great questions, every interaction helps make Coder better.</p>

<p>This is not just a development tool. It is a collaborative space where builders support builders, and everyone has something to offer. Whether you want to:</p>

<ul>
<li>Maintain or contribute to open source projects</li>
<li>Create tutorials, templates, or guides for others</li>
<li>Join discussions and help answer questions</li>
<li>Learn by exploring how others build their environments</li>
</ul>

<p>We invite you to be part of it.</p>


<h2>Here’s how to get involved</h2>

<h3><a href="https://discord.com/invite/coder">Join our Discord</a></h3>

<p>Chat with other developers, get help, share feedback, and stay updated on community events and announcements.</p>

<h3><a href="https://registry.coder.com/">Contribute to the Coder Registry</a></h3>

<p>Build and share your own workspace templates, or explore community-contributed ones to speed up your setup.</p>

<h3><a href="https://github.com/coder/coder/discussions">Join GitHub Discussions</a></h3>

<p>Share ideas, ask questions, and help shape the direction of our open source projects.</p>

<h3><a href="https://www.meetup.com/pro/coder/">Attend a Coder Meetup</a></h3>

<p>Meet other community members, learn from talks and demos, and connect in real life.</p>


<p>Whether you are here to give, learn, build, or all three, this community is yours. Help us shape what is next.</p>


<h2>Introducing the Coder Champions Program</h2>

<p>Coder continues to grow thanks to a passionate community that brings creativity, insight, and energy to everything we do. As we mark 100,000 GitHub stars, we want to recognize the people who make Coder more than just a product. It&#39;s a movement.</p>

<h3>What is it?</h3>

<p>The Coder Champions Program is the start of a broader initiative to recognize the builders, helpers, and advocates who go above and beyond. Whether they are mentoring others, building tools, sparking ideas, or simply lending a hand, these individuals make the Coder community stronger.</p>

<p>We are beginning with our first tier, <strong>Founding Coder Champions</strong>, which sets the stage for deeper involvement, meaningful recognition, and exciting rewards.</p>

<h3>How to qualify</h3>

<ul>
<li>Participate in five community <a href="https://discord.com/invite/coder">Discord chats</a></li>
<li>Star the <a href="https://github.com/coder">Coder GitHub repo</a></li>
<li>Help at least two people in the Discord #help channel</li>
</ul>

<h3>Perks of being a Founding Coder Champion</h3>

<ul>
<li>A unique badge on Discord</li>
<li>An exclusive T-shirt and sticker pack</li>
<li>Invite to a private voice call with the Coder team</li>
<li>Early access to our upcoming roadmap</li>
</ul>

<p>Nominate yourself or someone you admire by dropping a 🎉 on our Discord announcement. Once you meet the criteria, you are officially in. We also want to hear your feedback and suggestions to help shape the future of the program.</p>


<h2>Try out Coder today</h2>

<p>We released Coder as open source because we understand that developers look for transparency, a vibrant community, and low-risk evaluation when adopting a new platform.</p>

<p>Whether you are just discovering Coder for the first time or coming back for a fresh look, there has never been a better moment to get hands-on.</p>

<h3>Getting started is simple and flexible</h3>

<ul>
<li>Clone our GitHub repository and follow the quickstart guide</li>
<li>Deploy on your own infrastructure in minutes with full control and zero friction</li>
<li>Use prebuilt templates or craft your own to match the exact workflows your team needs</li>
</ul>

<p>Coder is built for developers who want speed without sacrificing security, power without giving up control, and modern workflows without vendor lock-in.</p>

<p>You can start exploring in under 10 minutes and make it your own in far less time than you would expect.</p>

<p>So go ahead. Launch your first workspace, customize it, break it, rebuild it, and see what Coder can do. Let’s keep building the future of development together.</p>


Recognizing Coder's developer community and the contributions they have made to our open source projects

100K stars on Github

100,000 Stars and Counting: Celebrating Coder’s Community and Commitment to Open Source

<p>Coder Community is the free-to-use tier based on our open-source release of Coder. Coder is a <a href="https://coder.com/cde">cloud development environment (CDE)</a>, it provisions and manages workspaces. Coder’s mission is to <a href="https://coder.com/solutions/productivity">onboard developers</a> to a productive, ready-to-use, fully configured cloud development environment. We released Coder as open source because we understand that developers look for transparency, a vibrant community, and low-risk evaluation when adopting a new platform.</p>


<h2>The Coder Community project</h2>

<p>As developers ourselves, we appreciate the value of working with transparently-developed software that includes an active community outside of its corporate maintainer. So in 2022, Coder released Coder as an <a href="https://github.com/coder/coder">open source project</a> under the <a href="https://www.gnu.org/licenses/agpl-3.0.html">GNU Affero General Public License</a>. </p>

<p>Since then, we’ve earned almost 9000 stars for this project in <a href="https://github.com/coder/coder">GitHub</a>. Small and large developer teams have discovered us and set up their own trial on their existing infrastructure. With Coder Community, there’s no need to contact a sales person for a license key or sign up for trial subscription. </p>

<p>We’re proud of the participation from like-minded developers that has grown for Coder. People can audit the code themselves to identify bugs and submit enhancements with <a href="https://github.com/coder/coder/issues">issues</a>. And they can join in on discussions on <a href="https://github.com/coder/coder/discussions">GitHub</a> and <a href="https://discord.com/invite/coder">Discord</a>. </p>

<p>We’re also proud of the substantial ecosystem of related projects. Coder hosts over 100 other open source repos with over 90,000 stars so far on GitHub. These tools support Coder and give back to the community. Our most popular repo by far is <a href="https://github.com/coder/code-server">code-server</a>, which runs <a href="https://github.com/Microsoft/vscode">VS Code</a> in the browser on any machine anywhere.</p>


<h2>Made by developers to solve developer problems</h2>

<p>Coder was started by developers, and is still maintained by many of the same developers, to solve a developer&#39;s problem: provide a great experience by applying the cloud’s flexibility, scalability, and stability to development environments.</p>

<p>Even a small development team runs into “<a href="https://coder.com/blog/it-works-on-my-machine-explained">works on my machine</a>”. Achieving a consistent dev environment for each developer means manually documenting the setup process. But this documentation only works when it’s actively maintained and everyone follows it consistently.</p>

<p>Large enterprise teams, on the other hand, soon learn that traditional methods such as local development, <a href="https://coder.com/solutions/vdi">VDI</a>, and shared VMs are less than ideal. For example, VDI uses expensive compute resources, including OS licensing, and suffers from a laggy local experience when working through a remote desktop client.</p>

<p>We made Coder to solve these and bigger problems for developers. Cloud technologies offer proven solutions to add flexibility, performance, scalability, and stability to software. Coder brings those benefits to software development itself: </p>

<ul>
<li><strong>Developers:</strong> Click a button to get coding in seconds with a ready-to-use development environment in the cloud. Developers work from their preferred local IDE, editor, or even just ssh.</li>
<li><strong>Enterprise:</strong> Deploy, monitor, and manage dev environments in the cloud, will full control over security and cost, from a consistent, cohesive, single place.</li>
<li><strong>Self-hosted:</strong> Coder’s flexibility includes choices for the infrastructure to deploy it on. You can install it in minutes on your laptop and you can scale it on a hyperscaler for hundreds of geographically-dispersed developers. Either way it’s the same Coder.</li>
<li><strong>Cloud-native:</strong> Coder is designed and built from the ground up for the cloud. It doesn’t depend on proprietary third-party software. It uses and works with established, proven cloud and open source technologies. These includes <a href="https://www.terraform.io/">Terraform</a>, <a href="https://www.docker.com/">Docker</a>, <a href="https://kubernetes.io/">Kubernetes</a>,  <a href="https://tailscale.com/">Tailscale</a>, OAuth, <a href="https://git-scm.com/">git</a>, and ssh.</li>
</ul>


<h2>Coder is great for small teams</h2>

<p>Startups, lone developers, hobbyists, and educators use Coder to give themselves the convenience and simplicity of starting up their development environments consistently and quickly.</p>

<p>Set up for small teams or even a single person is fast. You can install Coder on macOS, Windows, or Linux, on your laptop or your homelab, and get coding in minutes. You don’t need expertise in deploying cloud services and interconnecting them.</p>

<p>Coder Community fits a small team’s budget. There are no licensing fees, no subscriptions. As for hardware, even a modest laptop can run an IDE locally with snappy performance. To host the workspaces, use your on-prem hardware. Or, if you don’t have the budget for expensive hardware on-premises, then run workspaces in the cloud and pay for it only when the developer is at the keyboard.</p>

<p>Developers don’t need to install anything besides a browser and their preferred IDE. Or if they choose, they don’t even need an IDE. With <a href="https://github.com/coder/code-server">code-server</a>, developers can work from a browser-based IDE on a laptop or even a tablet.</p>


<h2>Coder is for the enterprise too</h2>

<p>Coder Community is for the enterprise, too. CDE adoption is a big undertaking for hundreds or thousands of developers. </p>

<p>First, there&#39;s no cost to trying out Coder Community. Start with a small pilot project then safely, effectively iterate toward a more mature adoption that really takes advantage of cloud-based development.</p>

<p>That means that there’s no initial hurdle of getting budget approval for licensing Coder or any other proprietary software. Trying it out can be as simple and straightforward as finding some spare compute and storage in your data center then installing Coder. </p>

<p>Deploying and administering Coder is straightforward too. There are no new or obscure systems to learn. Your platform team works with what they know, and onboarding new platform engineers is faster. </p>

<p>Most likely, your platform engineering team has security requirements. They can confirm their security concerns by checking out the <a href="https://github.com/coder/coder">Coder repo</a>, it’s publicly available. They can also follow the progress of issues, pull requests, and discussions.</p>

<p>When you reach the point where you need Coder&#39;s enterprise features, like RBAC or specific customizations, we&#39;re ready to help with <a href="https://coder.com/pricing">Coder Premium</a>, which is based on the same code as Coder Community.</p>


<h2>Try it out now</h2>

<p>We’ve made it as easy as possible to deploy Coder. Here’s a quick way to try it out now. </p>

<p>What you need:</p>

<ul>
<li>A spare macOS, Windows, or Linux machine to host Coder.</li>
<li>Optional: Another machine to develop from. You can also use the same machine that hosts Coder.</li>
<li>A few minutes.</li>
</ul>

<h3>Step 1: Install Coder</h3>

<p>Follow the <a href="https://coder.com/docs/tutorials/quickstart">Quick Start section</a>. It asks you to install <a href="https://www.docker.com/">Docker</a> which you’ll use to run workspaces on the same machine that hosts Coder. You’ll also create a new template from the Docker <a href="https://coder.com/docs/admin/templates/creating-templates">starter template</a>. Coder works with any other runtime platform, local or remote, but Docker is easy to set up.</p>

<h3>Step 2: Install an IDE</h3>

<p>Install the <a href="https://code.visualstudio.com/">VS Code</a> IDE on the local machine you’ll be developing on. Coder’s starter templates enable VS Code by default. You can use the IDE on the same machine as Coder, or on any other machine that has a network connection. </p>

<p>Coder handles secure networking for you so there’s no need to modify network configuration to connect a local IDE to a remote instance of Coder.</p>

<h3>Step 3: Start coding!</h3>

<p>During your Coder installation, you got a unique URL to your new Coder instance. In your web browser, sign in (if you aren’t already).</p>

<p>In the Coder web UI, select the <strong>Templates</strong> tab, select your new template, then select <strong>Create Workspace</strong>.</p>


Create a workspace from a template in Coder's cloud development environment

<p>Fill out the form to name your new workspace then select <strong>Create Workspace</strong>. It takes a minute or two to start a workspace for the first time because Coder needs to build the environment from scratch. After that, starting the same workspace takes just a few seconds.</p>

<p>When the workspace is running, you’ll see its page in the <strong>Workspaces</strong> tab. To connect VS Code to your workspace, just select <strong>VS Code Desktop</strong>.</p>


To connect VS Code to your workspace in Coder's cloud development environment, just select VS Code Desktop.

<p>The workspace has other choices to connect to it, including <a href="https://github.com/coder/code-server">code-server</a>, which is a great solution for small teams to work on limited workstations like tablets. And it’s great solution for enterprises to provide dev environments to contractors without needing to supply expensive laptops.</p>

<p>When you’re done with your workspace, select <strong>Stop</strong> in the workspace’s web page. Coder tears down the workspace for you. You can return to it any time by going to the <strong>Workspaces</strong> tab.</p>


<h2>We make Coder Community for you</h2>

<p>Coder gives developer teams a great experience by taking advantage of the cloud’s flexibility, scalability, and reliability. We made Coder Community an open source project to make it simple for you to try, inspect, deploy, and contribute to.</p>

<ul>
<li>Learn more about Coder, including installing it and setting up templates: <a href="https://coder.com/docs">Coder Docs</a></li>
<li>Check out the code: <a href="https://github.com/coder/coder">Coder repo</a></li>
<li>Check out related projects that Coder hosts: <a href="https://github.com/coder">Coder on GitHub</a></li>
<li>Browse templates and modules to extend Coder workspaces: <a href="https://registry.coder.com/">Coder Registry</a></li>
<li>Connect with the Coder community: <a href="https://discord.com/invite/coder">Discord</a>, <a href="https://github.com/coder/coder/discussions">Discussions</a></li>
<li>Subscribe to Coder&#39;s newsletter below</li>
</ul>


Coder Community is the free-to-use tier based on our open-source release of Coder. Coder is a cloud development environment (CDE), it provisions and manages workspaces so developers can quickly set up their development environments. Coder Community promotes transparency, community involvement, and is easy to deploy, making it suitable for both small teams and enterprises.

Coder Community: Open-Source Cloud Development Environment

Coder Community: Our Open-Source Offering

<p>We recently hosted a webinar exploring some recent movements in internal developer platforms (IDPs) and cloud development environments (CDEs). Our Head of Product, Ben Potter, demo’d the Coder Backstage Plugin to show how enterprises can improve developer experience while managing the complexity of the typical enterprise development environment. </p>

<p>Here’s what Ben covered: </p>

<ul>
<li>The complexity of modern development projects</li>
<li>How Backstage unifies all the items in a project</li>
<li>How Coder simplifies environment configuration and provisioning</li>
<li>How Coder and Backstage complement each other</li>
</ul>


<h2>Development projects are complex</h2>

<p>Developers in enterprise organizations wrangle a long list of services, APIs, and documentation for things like authentication, CI/CD, debugging, testing, orchestration, and so on. Keeping track of all these items in a project consumes a lot of developer time. Trying to use them effectively and consistently makes a developer’s life even harder.</p>

<h2>Unify with Backstage</h2>

<p>In the webinar, Ben gives an overview of <a href="https://backstage.io/">Backstage</a>, an internal developer platform. It provides organizations with a developer portal, which is a centralized knowledge hub for distraction-free, streamlined workflow.</p>

<p>Backstage’s <a href="https://backstage.io/plugins">extensive plugin ecosystem</a> lets organizations unify all of a project’s parts in a service catalog. This lightens the cognitive load on developers, giving them quick access to relevant resources.</p>

<h2>Now add Coder</h2>

<p>With Backstage’s unified service catalog, developers need a way to manage development environments. Coder does exactly this. Ben explains how Coder abstracts the complexities of securely and efficiently configuring and provisioning cloud development environments. Developers can start coding at the literal click of a button.</p>

<p>With a single source of truth for configuration, Coder lets organizations eliminate the discrepancies between development and production environments. And Coder&#39;s built-in security lets organizations protect sensitive code and data in the cloud-centric development landscape.</p>


<p>In the webinar, you’ll see  a practical example of how a couple of plugins from Coder bridge the gap between configuring their development environments and the coding process itself. Backstage and Coder bring cohesiveness to software development workflows, for an efficient approach to managing the software lifecycle. They lay a smooth path from project initiation to code development. </p>


<p>Developers can use Backstage to access project-related information and resources. Then Coder ensures that their development environment is consistent, secure, and tailored to the project&#39;s needs. This lets developers focus more on development and less on the logistical challenges that hinder the development process.</p>

<h2>Learn more</h2>

<p>If you want to dive deeper into using Coder’s Backstage Plugin, <a href="https://coder.com/webinars/backstage-coder-the-full-software-development-lifecycle/register">click here to register</a> and watch the full webinar.</p>

<p>Also, you can learn more about the <a href="https://coder.com/solutions/backstage">Coder Backstage Plugin</a> itself.</p>

<h2>Related Content</h2>

<ul>
<li><a href="https://coder.com/blog/coder-backstage-plugin">Introducing the Coder Backstage Plugin</a></li>
<li><a href="https://coder.com/blog/we-re-building-a-backstage-io-plugin">We’re building a Backstage.io plugin</a></li>
<li><a href="https://coder.com/solutions/backstage">Discover, Click, Commit: Coder + Backstage</a></li>
</ul>



Explore the intersection of internal developer platforms and cloud development environments in our recent webinar, "Backstage + Coder: The Full Software Development Lifecycle." Ben Potter, our Head of Product, demonstrates the powerful capabilities of the Coder Backstage Plugin, showing how it can enhance the developer experience and manage the complexities of enterprise development projects.

Backstage + Coder: The Full Software Development Lifecycle Webinar Recap

<p>Choosing a cloud-based development environment shouldn’t be hard. That&#39;s why we&#39;ve put together a streamlined overview of our two popular options: code-server and Coder. Both are web-based, secure and offer impressive features. Let&#39;s begin by exploring code-server, the project that paved the way for web-based coding environments.</p>


<h2>code-server (The one that started it all)</h2>

<p>code-server is an open-source project that transforms <a href="https://coder.com/blog/access-your-vs-code-environment-from-any-device">Visual Studio Code (VS Code)</a> into a server, enabling it to run on any machine. As a result, developers can access VS Code from any device with an internet connection, opening up new possibilities for remote development.</p>

<p><a href="https://coder.com/docs/code-server/latest">code-server</a> stands out for its compatibility with most VS Code extensions, thanks in part to its use of Open VSX, an open-source alternative to the <a href="https://marketplace.visualstudio.com/">Visual Studio Marketplace</a>. This feature allows developers to tailor their coding environment to their specific needs, even when working remotely. code-server customers can optionally host their own VS Code extension marketplace if they’re deploying Coder in a place without public network access e.g. corporate VPN. <em>Details can be found</em> <a href="https://github.com/coder/code-marketplace">here</a>. </p>

<p>One of code-server&#39;s significant advantages is maintaining a consistent development environment across diverse devices, reducing issues with code behaving differently on various machines. A user can connect from any device, as the IDE <a href="https://coder.com/docs/v1/latest/workspaces/editors">(Integrated development environment)</a> runs on a remote server. Yet, each code-server represents only one environment. </p>

<p>If a user needs to manage multiple environments or accommodate multiple users, they have two options, A) Launch multiple instances of code-server, potentially leading to inconsistency or configuration drift or B) Deploy Coder, a streamlined solution for maintaining uniformity and control. We will go more into Coder in a few moments.</p>


<h2>Highlights of code-server</h2>

<ul>
<li>Code on any device with a consistent development environment</li>
<li>Use cloud servers to speed up tests, compilations, downloads, and more</li>
<li>Preserve battery life when you&#39;re on the go; all intensive tasks run on your server</li>
<li>VS Code features are supported but not all extensions</li>
</ul>

<p><em>Note: For requirements to run code-server, please read our docs <a href="https://coder.com/docs/code-server/latest#highlights">here</a>. Code-server is a component of Coder. Think of code-server as a container and Coder as Kubernetes. More details about Coder will be provided below.</em></p>

<h2>Coder (The one built for teams &amp; enterprises)</h2>

<p>Coder is an open-source platform that enables the creation and management of developer workspaces on preferred clouds and servers. By leveraging common development interfaces (SSH) and infrastructure tools (Terraform), <a href="https://coder.com/docs/v2/latest/enterprise">Coder</a> simplifies the process of provisioning and accessing remote workspaces for organizations of varying sizes and stages of cloud-native maturity.</p>

<p>Terraform is used to represent Coder workspaces, though no prior <a href="https://developer.hashicorp.com/terraform/intro">Terraform</a> knowledge is necessary to use the platform. Coder provides a database of pre-made templates to facilitate workspace creation. A key benefit of using Terraform means that a workspace could be a VM or a container and run on any cloud or on-prem.</p>

<p>With Coder, workspaces are not limited to computing resources; users can add storage buckets, secrets, sidecars, and more using Terraform. Coder is free to use, but some features require a paid license. A trial can be requested at <a href="https://coder.com/trial">https://coder.com/trial</a>.</p>


<h2>Highlights of Coder</h2>

<ul>
<li><strong>On-Demand Development Environments</strong>: Coder provides remote development environments that are suitable for data and AI operations, allowing data engineers and scientists to access their favorite IDEs and robust server compute with ease.</li>
<li><strong>Enhanced IDE Support</strong>: While code-server offers a robust option for development, it&#39;s not the only choice with Coder. Developers have the flexibility to use familiar desktop IDEs from their local environments, like VS Code and JetBrains Gateway. Web IDEs such as Jupyter and RStudio are also available. Interestingly, some workspaces, particularly those focusing on data science, may choose to forego VS Code entirely, showcasing the adaptability of these platforms</li>
<li><strong>Dedicated and Secure Workspace</strong>: Coder&#39;s isolated workspaces can come with dedicated Python kernels and server-tiered resources, providing governance and control within a secure cloud infrastructure. Kubernetes and AWS security groups can be leveraged for workspace isolation.</li>
<li><strong>Cost-Effective and Efficient Computing</strong>: Coder ensures efficient cloud adoption by supporting all cloud machine types, automatically managing resource consumption, and eliminating financial barriers to cloud adoption.</li>
<li><strong>Improved Developer Experience</strong>: Users are not limited to web browsers. With Coder, users can utilize desktop IDEs, complete with their preferred configurations and shortcuts. Plus, the option to employ parameters and dotfiles for customization allows you to tailor your workspaces exactly to your needs, creating a truly personalized development environment</li>
</ul>


<h2>A brief demonstration of Coder just for You!</h2>


Short-Form Demo of Coder

<h2>Comparison table overview of code-server and Coder</h2>

<p>To make it easier to understand, here&#39;s a side-by-side comparison that highlights the differences between code-server and Coder.</p>


<h2>Which one is right for you?</h2>

<p>Both code-server and Coder are free and secure options to consider. If you are an individual starting out in remote development and need access to your desktop or remote server from a cafe, park, or anywhere with internet, use code-server. On the other hand, if you&#39;re looking for enhanced IDE support, on-demand development environments, a dedicated and secure workspace, cost-effective and efficient computing, and an improved developer experience, then Coder may be the right choice for you. Either way it’s a win win for your remote development future!</p>

<h2>Start a 30-day trial of Coder Enterprise</h2>

<p>If you would like to try our premium version of Coder, which includes all of the collaboration, security, governance, and management features, please visit <a href="https://coder.com/trial">https://coder.com/trial</a> to sign up.</p>


Choosing a cloud-based development environment shouldn’t be hard. That's why we've put together a streamlined overview of our two popular options: code-server and Coder. Both are web-based, secure and offer impressive features. Let's begin by exploring code-server, the project that paved the way for web-based coding environments.

AI-native Development

Secure Environments for 2,500+ Developers

Marco Martinez

Explore the latest from Marco

Building Together: Coder's Community Vision for 2026

How I Used AI to Solve a Real Community Scaling Problem

Vibe Coding My First App with Blink (No CS Degree Needed)

The AI Learning Curve: Why Devs Get Slower Before They Get Faster

Why Your AI Agents Need Ephemeral Environments (And Your Devs Probably Do Too)

100,000 Stars and Counting: Celebrating Coder’s Community and Commitment to Open Source

Coder Community: Our Open-Source Offering

Backstage + Coder: The Full Software Development Lifecycle Webinar Recap

Related content

Using Claude Code with Coder Workspaces

Introducing Coder Agents

Self-Hosted Doesn’t Always Mean What It Implies