Building @cognition. Previously @scale_AI @MIT
Joined October 2016
- Tweets161
- Following192
- Followers5.9K
- Likes608
steven hao retweeted
I was shocked to wake up and see *language server support*! Makes cloud dev feel so much closer to home
(this might only be available if you use both Devin and Devin Review)
@ everyone who works at a company looking for answers to "what should we do about security", here's an excellent answer. I'd say this is the 80/20.
Most people trying to secure their systems don't understand the threat model -- who and what are you defending against? Most exploit chains in practice are just variations of phishing/supply chain attack and use local employee devices as the break-in point. It's hard to categorically eliminate malware on local devices, but it's easy to limit the blast radius and ensure that a single compromised device doesn't end your entire company.
If you could pick 1 thing to do: Require phishing-resistant MFA (like a yubikey) before employees can access any data in production. 2FA is simple but unreasonably effective, and extremely underutilized. Passkeys, TouchId/FaceID, and AI tools like Devin make it really easy to implement it too.
1. Mercor has insane revenue growth
2. Mercor is a heavy user of Devin
This is not a coincidence! We're seeing more and more amazing tech companies adopt Devin for all their work and seeing huge efficiency gains as a result.
Founded by three 22-year-olds in 2023, @mercor crossed $2B ARR in June, only 4 months after hitting $1B.
Today, millions of experts on the platform bring their knowledge of medicine, law, engineering, and beyond to frontier AI - helping models take on increasingly complex and consequential work in the real world.
Mercor's engineers move at extraordinary speed.
They give Devin long-running tasks before bed, during meetings, or on their way to the gym. They come back to tested features and PRs ready to merge.
We've been sprinting to launch this new improved Devin Security -- now powered by frontier cyber models.
We're offering free scans to eligible enterprises starting TODAY, because when it comes to security, every second counts.
steven hao retweeted
Devin Fusion has claimed a top score of 68 on the @ArtificialAnlys Coding Agent Index with an Opus 5.5 lead and GPT-6 Luna sidekick.
Devin Fusion outperforms Claude Code with Opus 5.5 on both performance and price – earning a score of 68 at a 31% lower cost per task.
steven hao retweeted
There are real safety and security issues raised by AI, but they can be fixed. It's time to get to work instead of freaking out. That's why I'm joining Cognition as CISO.
steven hao retweeted
Devin now has a Mac
Here are 11 experiments I've run over the past week, as well as open source code for over 100 more native apps.
Build, test, deploy iOS, iPad, and macOS apps in the cloud with just a prompt. ⚡
1. Build and test a multi-platform, multiplayer app in the cloud (iOS, iPad, web)
steven hao retweeted
8. Multiplayer, multi-platform Chess.
Building and testing in the cloud with Devin computer use.
steven hao retweeted
I just tried SWE-2.0 on Devin desktop,& wow😮 it’s extremely good at coding!
At high effort, SWE-2 is definitely better than GPT-5.6 Sol high; at max level it is as good as GPT-6 Astra medium at less than half the cost! This is definitely a frontier-level model, just amazing!
steven hao retweeted
We independently benchmarked Devin Fusion for its release today - this is the first time a multi-model coding agent has been included on the Artificial Analysis Coding Agent Index, and it effectively retains Claude Fable 5.1 and GPT-6 Astra performance while reducing costs
Devin Fusion runs a frontier lead model with a cost-efficient sidekick. We tested configurations from Cognition combining frontier models from Anthropic and OpenAI with their new SWE-2 (medium) as a sidekick model. Configured with Claude Fable 5.1 (xhigh) + SWE-2 (medium), Devin Fusion scores 62 on the Coding Agent Index v1.5, while with GPT-6 Astra (xhigh) + SWE-2 (medium) it scores 59. The Fable configuration has the higher score, while the Astra configuration is 43% less expensive and completes tasks 31% faster.
Congratulations to @cognition on the release! See below for our results and analysis 🧵
steven hao retweeted
Introducing Devin Voice 🦦 ☎️
Your favorite AI software engineer just got a landline.
You say it, Devin ships it.
Powered by GPT-Live and our new SWE-2 model.
The dioxus team is insanely cracked. As a teaser, there's some features coming up in the next few weeks that will unlock significant new ways of using Devin.
These would not have been possible without the @jkelleyrtp and the rest of the Dioxus team
Welcome @jkelleyrtp and the entire Dioxus team to Cognition!
Dioxus is one of the most beloved open source Rust frameworks. We're proud to continue supporting Dioxus, Blitz, Taffy, and Subsecond while bringing the team's expertise to Devin's VM, computer use, and testing.
Our best model yet! The best option if you want something smart yet cost effective.
steven hao retweeted
Introducing SWE-2, our closest model yet to the frontier.
On leading evals, it scores on par with recent frontier models – at up to 70% lower cost.
We scaled RL to multiple trillions of parameters, with a refined recipe that pushes the Pareto curve on both capabilities & cost.
steven hao retweeted
Devin is the best software product I have ever used hands down
We are:
- 2 veteran engineers (inc me) and 1 PM
- 100% code is generated by Devin
- 7x the pace of pre Devin
- $8M ARR now
: )
steven hao retweeted
The world needs far more software than it can build. Cognition exists to change that.
We’ve just raised over $2B at a $48B valuation, led by a16z, Accel, Founders Fund, General Catalyst, and Avenir.
Since our round in May, run-rate revenue has grown from $492 M to almost $900 M.
