Jeff Wang @jeffwang

President, New Enterprise at Cognition, Prev. Windsurf CEO, AI @ https://t.co/AxabQu0R8E San Francisco, CA Joined February 2009

Tweets

2K
Followers

13K
Following

372
Likes

1K

Devin Desktop @devindesktop

a day ago

Inkling and Grok 4.5 are also available in Devin Desktop and CLI

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.

68 81 728 1.5M 138

9 11 129 11K 10

View Details

Jeff Wang @jeffwang

2 days ago

Chinese open source is no longer "6 months behind", but it's also no longer "10% of the cost" either Congrats to Moonshot on Kimi K3, 2.8T, $3/$15 - same as Sonnet but initial feedback is great, we'll be sure to run FrontierCode on our side

60 129 4K 170K 209

View Details

Jeff Wang @jeffwang

4 days ago

See our joint post here: cognition.com/blog/one-year-…

0 0 18 13K 9

View Details

Jeff Wang @jeffwang

4 days ago

Exactly one year ago was the craziest 72 hours I’d experience in Silicon Valley. I was thrust into a situation where many members of the Windsurf team had moved on to Google, recruiters were reaching out to every member of the team, customers were calling to understand what was happening. It was definitely the scariest time of my life, and I was caught in a blur of nonstop phone calls and meetings to figure out what to do next But time seemed to slow down when me and Graham met with Scott and Russell. Everything seemed to click and make sense. They were looking for a GTM org, and we were looking for an Engineering org, both of which had been considered the best in Silicon Valley. Our products also did not overlap, we had been trying to build a cloud agent, they had been trying to build an IDE. It almost seemed too perfect, and 72 hours later we didn’t sleep to get the acquisition done. Now the part that isn’t told as often is what happened next. Windsurf quickly released Wave 11 and then a slew of other new features. We had a massive bug bashing effort to really polish the experience. Over the last year this led to unifying the platform around Devin. Now the future is ahead of us to continue to shape. Windsurf was the first agentic IDE, Devin was the first cloud agent. We’re now seeing new form factors quickly emerge with our customers. For example event driven agents are agents that automatically take on tasks based on triggers, conversations, and general awareness about the company. Devin can also manage swarms of agents, each with their own computer to get work done, as an example our security swarm is one of the most popular new features across our enterprises. What happened a year ago didn’t slow us down, it’s made us go faster, and we’re looking forward to building the future of software engineering.

21 15 399 235K 151

View Details

Jeff Wang @jeffwang

5 days ago

This is a pretty eye opening chart. By using frontier models as task managers and a sidekick to carry out these tasks, you can get Fable level performance for even cheaper than Opus 4.8 by a lot (40%)

Joon Lee @joon_h_lee

5 days ago

x.com/i/article/2076…

41 88 942 717K 2K

11 10 142 29K 89

View Details

Jeff Wang @jeffwang

6 days ago

@mattbergland How thoughtmaxxing of you to point this out

0 0 1 236 0

View Details

Jeff Wang @jeffwang

a week ago

I am humbled every time I step into the office, but the folks at Cognition are also awesome and fun to work with as well

Chris Barber (in SF) @chrisbarber

a week ago

I asked people which companies have the highest density of talented people they know: 1st. Cognition - 9 votes Equal 1st. Anthropic - 9 votes 2nd. Modal - 5 votes 3rd. OpenAI - 4 votes 4th. Standard Intelligence - 3 votes 4th. Cursor - 3 votes Two votes each: - Ramp - Flapping

76 47 1K 903K 2K

0 0 56 5K 2

View Details

Jeff Wang @jeffwang

a week ago

@vincentweisser @swyx Congrats @vincentweisser!!

1 0 3 270 0

View Details

Jeff Wang @jeffwang

2 weeks ago

@tomasholtz @cognition Welcome to the team!

0 0 2 188 0

View Details

nader dabit @dabit3

2 weeks ago

Shipped a new iOS shortcut - kick off Devin Cloud Agents using your action button on iOS, or with one click from your home screen. You can also create sophisticated automations for any connected repo with natural language. Give it a shot! github.com/dabit3/devin-i…

21 11 146 20K 62

View Details

Jeff Wang @jeffwang

2 weeks ago

Security hardening is a big initiative at almost every company I talk to In keeping with the momentum of our Security Agent Swarm release, we’re now launching a Vulnerability Remediation program We want to auto remediate and burn down your vulnerability backlogs in weeks Check out our program here and get in touch devin.ai/security-progr…

2 2 22 1K 0

View Details

Jeff Wang @jeffwang

2 weeks ago

Security is still top of mind and we’ve released Devin Security Swarm Now you can proactively scan code bases with a team of agents, verify and consolidate findings, and of course patch them as well We do this for much cheaper than other alternatives while being even more comprehensive

Cognition @cognition

2 weeks ago

Introducing Devin Security Swarm A more cost effective and accurate way to find security vulnerabilities in complex codebases, based on a new architecture: Agentic MapReduce.

58 94 1K 671K 598

1 0 33 3K 2

View Details

Jeff Wang @jeffwang

3 weeks ago

We’ve developed Devin Fusion, an agent that utilizes a “sidekick” to give you Fable level performance but 35% cheaper This doesn’t come as a tradeoff to performance, we have utilized FrontierCode to bring this to you and will also continue to do so as new models come out

Cognition @cognition

3 weeks ago

Conventional model routing sucks. It passes benchmarks but fails to write code you'd actually merge. Introducing Devin Fusion, a new hybrid-model harness for agentic coding. In testing, it reduces the cost of Fable-level intelligence by 35% and still feels good to use.

80 130 1K 535K 538

7 5 76 8K 12

View Details

Jeff Wang @jeffwang

3 weeks ago

Wow incredible scores from GLM-5.2 and Kimi K2.7, ranking just behind GPT-5.5 and Opus 4.8 And just for fun, we'll make these free for a while for Devin Desktop and CLI, try these out today!

Devin Desktop @devindesktop

3 weeks ago

Kimi K2.7 Code and GLM 5.2 are available in Devin Desktop and CLI Both perform strongly on FrontierCode Extended, our benchmark for real-world engineering tasks: GLM 5.2: 43.0% Kimi K2.7 Code: 39.5% Pro/Max/Teams users can try both models for free until July 5

31 51 453 799K 113

4 0 58 4K 3

View Details

Jeff Wang @jeffwang

3 weeks ago

New problem, how do I avoid eating too much at @cognition

38 0 288 53K 33

View Details

Jeff Wang @jeffwang

3 weeks ago

By far the most underrated feature of Devin right now is actually proving the task is done correctly You can work with Devin on a comprehensive plan, then actually run the plan out and send back a checklist and video that it did the job correctly Combine this with robust infra to run operating systems and applications, and you can truly run parallelized work

Cognition @cognition

3 weeks ago

Devin tests its work before you review the PR. You review and approve the test plan. Then get back a screen recording with a visual checklist of step-by-step QA.