Our users have been really excited to try Kimi K3 and we're excited to bring it to Augment's product family. It is the most capable open-source model that we have tested to date. Try it today in Cosmos, our agent orchestration platform! Congrats to the Kimi (Moonshot AI), and thanks to Fireworks AI for offering us US-hosted inference from day one.
Augment Code
Software Development
Palo Alto, California 19,510 followers
More agents ≠ engineering transformation. Meet Cosmos, the OS for agentic software development.
About us
Augment is the AI coding platform built for enterprises - powered by a highly secure, production-grade Context Engine developed over 2.5 years in stealth until our public launch in November 2024. Augment semantically indexes your entire codebase, documentation, dependencies, and internal knowledge in real time to enable devs to ship code faster, onboard in days instead of months, and modernize legacy systems with confidence. We are alumni of global-leading AI and cloud companies including Google, Meta, NVIDIA, Snowflake, and Databricks - all on a mission to provide enterprises with context that actually understands your codebase. If, like us, you believe in augmenting and not replacing software developers, join us on our mission to improve software development at scale using AI.
- Website
-
http://www.augmentcode.com
External link for Augment Code
- Industry
- Software Development
- Company size
- 51-200 employees
- Headquarters
- Palo Alto, California
- Type
- Privately Held
- Founded
- 2022
- Specialties
- AI, Software Engineering, Developer Tools, Platform Engineering, and Developer Productivity
Locations
-
Primary
Get directions
Palo Alto, California, US
Employees at Augment Code
Updates
-
Opus 5 is live in Augment Code. In our internal evals it scored higher than Opus 4.8 on correctness, completeness, and code reuse. We trust it enough that we're upgrading one of our Cosmos code review experts that reviews all our code to Opus 5: near-Fable review quality at a lower price. Try it in Cosmos today!
-
-
Augment Code reposted this
One of the principles you learn early in software engineering: ship small changes, frequently. I don't think that matters in the AI era. DX's excellent report published this week on the impact of AI in engineering shows that everyone is moving away from incremental delivery. They frame that as a problem, but is it? I'd argue no. It's important for humans to ship small changes for a lot of reasons: momentum, rollback, cognitive overhead, quick review turnaround. But agents don't share those limitations, and in fact can be very efficient one-shotting large projects. Links to the DX report and how we build large projects at Augment Code below.
-
-
What is loop engineering, and how are teams building with loops? Boris Cherny mentioned that he doesn’t prompt Claude anymore, he writes loops. 🦄 Peter Steinberger posted a monthly reminder to stop prompting agents, and design loops that prompt them instead. On the other side, people are pointing out that a loop around an LLM call is still just a while loop, and that we had cron jobs and webhooks long before we had agents. I think they are both correct. There are loops at an individual level; and there is a new pattern where AI forward teams are building agentic SDLC loops. Underneath this debate, there is an important shift that’s happening: the work inside the loop has changed. Agents can now investigate, modify code, run tests, and review results with pretty high quality. The middle of the software development lifecycle (SDLC), where humans used to spend so much time working on repetitive tasks, can now be designed with agents and humans working side-by-side, as part of an autonomous system. I wrote about how some of the most forward leaning AI teams are building loops on Cosmos and where they are finding the most value. https://lnkd.in/gvfbYXrd
-
-
For the first time ever: Grok 4.5 is now available. Welcome SpaceXAI to Cosmos, our agent orchestration platform. Together with Augment's context engine, Grok 4.5 is a powerful option for working across large codebases. We're interested to see how Grok 4.5 performs for our customers - let us know your thoughts and what you plan to build.
-
What does the role of an engineer look like today? In the latest episode of We Built What? Wei-Wei W. and Emma Webb take a deep dive on the skills required. Full episode in the comments below 👇
-
GPT-5.6 Sol, Terra, and Luna are now available in the model picker! Try them today in Cosmos, our unified agent platform! If you do not see GPT 5.6 in your model picker, your Augment Code admin must first opt into the Responsible ZDR option in the admin dashboard, which is needed for models requiring AI Safety Programs.
-
New episode of We Built What? just dropped and its a good one 👇️
The most overhyped thing in AI testing? Wei-Wei W., Co-founder and CEO Momentic says its using production as your source of truth. At first I didn't get it: aren't we supposed to test in production? Then it clicked. The opportunity space isn't what exists today. It's the delta between what exists and your ideal. Testing against whats in production only makes sense if production is perfect, which is...never. Full episode of Augment Code's We Built What? in comments - thanks again for joining, Wei-Wei W., let's do it again soon!
-
Raise your hand if you have zero tickets in your backlog? 🙋♂️... 🦗 The backlog is where great ideas usually go to wait. But with the new Linear and Cosmos partnership, "in progress" now really means what it says. In demo linked in the comments, Phillip Booth shows you how to turn a Linear ticket into a full technical spec and a polished PR without leaving your workflow 👇
-
-
Augment Code reposted this
"Build the factory, don't work on the factory line." In this episode of Augment Code's We Built What? I talk to my friend, colleague, and 🐐 Chris Kelly about what it means to be an engineer right now, and why the job isn't "write more code faster." It's design the system that gets an LLM to produce the code you want. A few things that stuck with me: Chris has spent his career on deterministic software: same input, same output, every time. Cosmos asks him to give that up. Instead of writing scripts that run the same way forever, he's building agents that manage their own subscriptions, make their own calls, and course correct when they're wrong. The guardrails look different now, too: less "if this, then that", more verification loops between every handoff. He also thinks the IDE has run out of road, because work now starts somewhere else: a bug report, an incident, a Slack thread, a weekly digest an agent generated on its own. Humans are in the loop, but the loop no longer happens in the IDE. And then there's the adoption problem every engineering leader is living right now: one person on your team running 500k tokens a month, another running five. Chris's take is that you can't fix that with better prompting habits. You have to build the tooling so the whole org inherits the best solution the moment someone finds it, the same way open source culture has always worked. My favorite line from him: we're not trying to build a faster horse. We're trying to build a car. And figuring out how to drive it is going to be the hardest and most interesting part of the job for a long time. Give it a listen 🎧