@cognition
Makers of Devin, the first AI software engineer. We are an applied AI lab building end-to-end software agents. Join us: https://t.co/4Ss9hvpjRG
Get your slippers on and join us for a walk around our SF headquarters, where we're doing it all with Devin. Even ordering our morning coffee. cognition.com/careers
The Devin webapp is getting a facelift. 80% less loading lag, customizable sidebar, do everything with ⌘K, and more.
We rebuilt our chat renderer from scratch for faster load times and smooth scrolling, even in the longest sessions. Long chats open 55% faster and INP is down 36%. devin.ai/blog/rebuildin…
In Prague, Czechia, you can order a week's worth of farm-fresh groceries and have it at your door in an exact 15-minute window. Milk in glass bottles from a local family farm. Fish caught yesterday. Strawberries picked that morning. This is all fueled by @rohlikgroup, a powerhouse from Czechia. It’s profitable, it did >$1.3B of revenue last year, and it's accelerating European grocery delivery with Devin.
@rohlikgroup Read all about it: devin.ai/customers/rohl…
GPT-5.6 Sol is now the most affordable frontier model on Devin Desktop and CLI. Today, OpenAI reduced Sol’s prices by 20% for the next 3 months. Last week's 70% discount still applies, so through October 3 you now get a 76% discount off the list price.
Read more on our blog: devin.ai/blog/gpt-5-6-s…
Introducing @SlackHQ Code in Devin. Now, Devin proactively creates dedicated code channels to keep work focused. Read more about how we tuned Devin’s personality in Slack: devin.ai/blog/devins-sl…
@SlackHQ Proud to be a launch partner with @SlackHQ.
Slack@SlackHQ·It's time to be ambitious.
Scott Wu@ScottWu46·At Cognition, one of our values is to go for it all: when faced with a tradeoff, pick the ambition-maximizing direction. When we launched Devin in 2024, we envisioned a future where every team had an infinite army of junior engineers. We were early and Devin wasn't good enough. Now it is! The only constraint left is how ambitious you're willing to be.
Thanks for the great week, Tashkent! #ioi2026 Cognition is proud to sponsor this year's IOI and support the new generation of competitive programmers.
GE built America's first jet engine in 1942. Today GE Aerospace's engine technology powers 3 of 4 commercial flights in the world. Now they’re using Devin to build more, faster. One team nearly doubled its engineering output. Their software optimizes routes, reduces fuel consumption, and monitors anomalies. Their software team's goal is to fly more without adding another airplane, runway, or pilot.
GE built America's first jet engine in 1942. Today GE Aerospace's engine technology powers 3 of 4 commercial flights in the world. Their software team builds tools that optimize routes, reduces fuel consumption, and monitors anomalies. The goal is to help airlines fly more without adding another airplane, runway, or pilot. Now they’re using Devin to build more, faster. One team nearly doubled its engineering output. From America’s first jet engine to AI, GE Aerospace keeps inventing the future of flight. It's a privilege to support their work.
GE built America's first jet engine in 1942. Today GE Aerospace's engine technology powers 3 of 4 commercial flights in the world. Their software team builds tools that optimize routes, reduces fuel consumption, and monitors anomalies. The goal is to help airlines fly more without adding another airplane, runway, or pilot. Now they’re using Devin to build more, faster. One team nearly doubled its engineering output. From America’s first jet engine to AI, GE Aerospace keeps inventing the future of flight. It's a privilege to support their work.
The weights for the X algorithm have been publicly released. The highest boost to a post comes from copying its link to share it. (Try it on this one for a surprise!)
The worst interactions are: 3. Not interested (-43.2) 2. Mute author (-58.8) 1. Report post (-234.0)
Gemini 3.7 Flash is now available in Devin. On FrontierCode 1.1, it reaches Claude Sonnet 5-level performance at less than half the cost, with the low latency the Flash series is known for.
Within Devin, Gemini 3.7 Flash performs particularly well on tightly scoped refactors, where it delivers minimal diffs that match repo conventions. Gemini 3.7 Flash is available at a 50% discount for 2 weeks, through 08/27/2026. Read more: devin.ai/blog/gemini-37…
Devin now works directly on your live data in @MongoDB Atlas. Query it, manage it, provision a new cluster mid-task. No stale schemas, no copied-in context, no waiting on someone to set up a database. Get started today by provisioning a MongoDB database directly from Devin.
Congratulations to Cognition engineers @sampriti0 and Alex Lin for winning DEF CON CTF!
perfect blue@pb_ctf·We won DEF CON CTF! perfect blue first formed in 2018 for DEF CON quals. We went from not qualifying, to 6th, 5th, and 2nd place. Dreams do come true🥲 This was the last big CTF we wanted to win. It's bittersweet but a good end. Thanks for all the great memories over the years.
Grok 4.6 is now available in Devin. Grok 4.6 marks a significant improvement over Grok 4.5, surpassing GPT-5.6 Sol, behind only Opus 5 and Fable 5.
Within Devin, Grok 4.6 shows particular strength on thorough code exploration and root cause analysis before touching any code. Read more: devin.ai/blog/grok-4-6
Cloud agents work while you eat, sleep, and take the subway. "Every morning I'll spin up a session on the way to work." Melius is a small and nimble startup, and Devin gives them the engineering capacity of a 100-person team. That's why the fastest moving startups use Devin.
For ambitious startups, we're offering $65k in Devin credits as a part of our startup program. Apply today: devin.ai/startups
Run Devin Outposts on @Vercel Sandbox to build and test your app in an isolated microVM. In Vercel Sandbox, Devin can run Docker, connect to private networks through a VPN, and resume from a filesystem snapshot with its repository, dependencies, and build state intact. x.com/vercel_dev/sta…
Learn more about Devin Outposts: docs.devin.ai/cloud/outposts…
Welcome to Cognition, Art! x.com/artlevy/status…
Thanks to improvements in both the harness and models, Devin Fusion is now 4% more intelligent and 27% less expensive on FrontierCode 1.1.
Try Devin Fusion today in Devin Cloud: app.devin.ai/signup
Welcome Paul Grewal to Cognition. x.com/iampaulgrewal/…
Using Devin for native iOS development: x.com/dabit3/status/…
With Outposts, Devin can natively run and test apps on any computer. Learn more: devin.ai/blog/introduci…
We've updated FrontierCode 1.1 to reflect new discounts for GPT-5.6 Terra and GPT-5.6 Luna. With these new costs, the GPT-5.6 series sits on the pareto curve of price/performance efficiency.
Check out all model scores, methodology, and sample tasks: cognition.com/frontiercode
Devin now natively supports @Github Stacked PRs ▪︎ Break down large changes into smaller, reviewable diffs ▪︎ Address and fix comments across stacks ▪︎ Automatically rebase downstream changes
Learn more about Stacked PRs in Devin: devin.ai/blog/introduci…
We're proud to join @NVIDIA and the Open Secure AI Alliance. To support open source models, we're contributing our research on measuring the trustworthiness and security of open source models. Closing open source models hurts innovation. The path forward is better tools to evaluate, secure, and deploy them responsibly.
Read more about our trustworthiness eval, which tests whether models repeat propaganda, comply with problematic requests, or write less secure code depending on who they’re working for. Our results show these risks aren’t inherent to open models and can be substantially mitigated through careful post-training. t.co/Q7VRccpSYi
Kimi K3 is now available in Devin Desktop and CLI. On FrontierCode 1.1, Kimi K3 is the first open source model we tested that approaches frontier-level performance.
On FrontierCode 1.1 Extended, our benchmark for real-world engineering tasks that grades mergeability and quality, Kimi K3 scores 58.2% with a 63.6% pass rate. Within Devin, it excels on reproducing bugs and managing its environment effectively. devin.ai/blog/kimi-k3
Claude Opus 5 is now available in Devin. On FrontierCode 1.1, Opus 5 approaches Fable-level performance at half the cost.
On FrontierCode 1.1, our benchmark for real-world engineering tasks that grades mergeability and quality, Opus 5 scores 63.6% with a 69.6% pass rate on Extended — approaching Fable 5 at half the cost. Within Devin, it shows particular strength on difficult debugging and root-cause analysis tasks. t.co/LcqxXbrUkp
DeepWiki has indexed over 500,000 repos. It's an easy way to get started when working with unfamiliar libraries, and free to use. In this meetup with @LangChain and @jacobtpl, we'll share how it works under the hood. x.com/LangChain/stat…
Cognition is acquiring @Interaction, the makers of Poke. If you've ever used Poke, you know why we're so excited.
Read more: cognition.com/blog/interacti…
Introducing Devin Outposts: run Devin on any machine. Your Mac mini, a GPU box in your lab, a VM inside your private network, or a Kubernetes cluster next to your internal services.
Outposts lets you run Devin sessions inside infrastructure you control. Devin’s agent loop (inference and planning) continues to run in Devin’s cloud, while all command execution, file edits, and repository access happen on machines you operate. Set up Devin Outposts for Devin Cloud using our docs: t.co/i12aSxissm
We’re thrilled to welcome @anhangz and @yunpark93 from TierZero to Cognition! TierZero builds agents that keep your product running: handling incidents, alerts, customer issues, and reliability. Together, we’re bringing these capabilities and more to Devin Automations. x.com/anhangz/status…
Devin Automations is our platform for turning engineering workflows into autonomous agents. Anhang and Yun are exceptional founders and product builders. And at Cognition, we love working with founders. More on why we’re so excited: cognition.com/blog/welcoming…
The FrontierCode leaderboard is now live: a dedicated page that tracks which models are writing code you’d actually merge. All scores — including Grok 4.5 and Inkling — are available, along with full methodology and sample tasks.
See the scores: cognition.com/frontiercode
Introducing Devin for Startups: $65k in credits to use Devin across Cloud, Desktop, and CLI. Apply today at the link below
In addition to credits, startups accepted will get access to exclusive events, merch, and white glove support. We’re hosting our first event in two weeks here: luma.com/devin-for-star…
Ship straight from the conversation. Devin in @Slack lets your team investigate issues, answer codebase questions, and start dev tasks without leaving the channel. Thanks for the feature, @Slack👇️ x.com/SlackHQ/status…
One year ago today, Cognition acquired Windsurf. In the year since, we've shipped dozens of new features, added hundreds of millions of dollars in ARR, and published frontier AI research. @ScottWu46 and @jeffwang share what we've accomplished in one year of building together.
As of today, Devin Fusion now incorporates Fable 5. Surprisingly, Fable 5 runs at a lower cost per task than Opus 4.8. Even though it’s a more expensive model, we observed cost efficiency improvements in areas like delegation and reasoning chains.
We found Fable works more like a competent manager delegating to an engineer, by writing comprehensive specs, giving detailed feedback, and trusting its sidekick with execution. Other models like Opus 4.8 tend to micromanage: they redo work, reread referenced files, and stuff their context window.
Yesterday we launched SWE-1.7 built on the open-source Kimi K2.7. Concerns about Chinese base models are real: K2.7 completed 87% of tasks that other models refuse over human-rights concerns. We trained SWE-1.7 specifically on trustworthiness, so it matches US models on evals.
When conversing in Simplified Chinese, leading open source models generally exhibit higher propaganda rates, alignment with CCP narratives, and refusals on politically sensitive questions.
GPT-5.6 is now available in Devin! On FontierCode 1.1 Extended, the GPT-5.6 family stands out for pairing strong scores with excellent cost efficiency. GPT 5.6 Sol reaches top performance at nearly half of the cost of the next best model.
Try it now in Devin Cloud, Devin Desktop, and Devin CLI. Read more at devin.ai/blog/gpt-5-6
Introducing SWE-1.7, the most capable model we’ve trained yet. It scores within a few points of the strongest frontier models at a fraction of the cost, and is now available at 1000 tok/s. RL is not hitting its limit: after refining our recipe, we keep seeing gains as we scale
SWE-1.7 was built on broad improvements in RL pipeline on top of a Kimi K2.7 base model. Our proprietary benchmark FrontierCode evaluates whether a model makes code you’d actually want to merge. SWE-1.7 advances the cost-performance Pareto curve on FrontierCode with a score of 42.3% and cost per task of $1.97 on our Main set.
We’ve made improvements to the FrontierCode methodology and are releasing FrontierCode 1.1 with clearer guidelines for fair internet use and refined grading criteria. Learn more about the changes on our blog: cognition.com/blog/frontier-…
Research, design, and math are three fields that tend to attract obsessives. Few have had a bigger impact on our imagination than legendary creator Grant Sanderson of @3blue1brown. We were lucky to host Grant at our office last week with Cognition's Head of Research @silasalberti and head of Design @josephhhhz for a fireside chat on the craft of making abstract ideas accessible, and how intelligence and design will evolve together. Events like this happen regularly at Cognition - if that sounds interesting you, you might be a great fit for our team. Check out our open careers here: t.co/ZsKXnw6E2f
Our goal is to help security teams shift from reactive triage to proactively fixing what matters, close the critical gaps scanners miss before they become breaches, and return engineering capacity to the product roadmap. Learn more and enroll today: devin.ai/security-progr…
Read the announcement: cognition.com/blog/devin-sec…
