Proximal builds infrastructure that enables AI to improve from real-world experience. In the past year, we have grown to more than $200M in annualized revenue helping frontier labs and enterprises improve coding agents. Now, we are expanding beyond software engineering. Recent progress in mathematics demonstrates what happens when models are trained in domains with massive amounts of public data and easy verifiability, making it easy to find weaknesses, generate training tasks, and iterate quickly. We build infrastructure that enables this feedback loop in other domains: we are excited about a world in which AI systems design targeted drugs, accelerate chip design, and rewrite legacy software that still runs hospitals, power grids, governments, and other critical infrastructure. Since starting, we have assembled a small team of researchers and engineers and built world-class infrastructure for post-training and synthetic data research to overcome the challenges of scaling manual data curation. Our team comes from Cursor, Google DeepMind, Meta Superintelligence, Prime Intellect, Citadel, and Jane Street. More than half are former founders. We’re fortunate to be backed by investors who share that vision, including @generalcatalyst who led our $15M Seed Round at a $300M valuation, as well as @chemistry, @svangel, @dvlamoen and individuals like @LiamFedus, @kevinweil, and @bernhardsson

Excited to share i’ve joined @ProximalHQ here in SF! I think that there is a ton of interesting open problems related to data and post training at large. There is no better team to work with than the one we have and I am extremely excited to share our work :)

Building robust benchmarks and training tasks in any domain is very difficult In cybersecurity benchmarks that purely rely on differential execution to understand the validity of a PoC, we observe a high rate of false positives and negatives Great work by @_anishlk!

Many cybersecurity evals use differential execution to test whether agents can reproduce known software vulnerabilities Without added constraints, this creates fairness issues: Our evaluation shows that CyberGym is effectively saturated on a verified task subset

Introducing Gemini 4 Argon – our new frontier model. It’s built for complex workflows across coding, enterprise knowledge work, and cybersecurity defense – rolling out today to a set of trusted testers through our Fairwind Program.

Introducing Gemini 4 Argon, our new frontier model, rolling out to cyber defenders starting today, and more widely as soon as possible. I am really excited by the progress we have made here. Argon is priced at $2 in and $10 out during introductory pricing!

Proximal builds infrastructure that enables AI to improve from real-world experience. In the past year, we have grown to more than $200M in annualized revenue helping frontier labs and enterprises improve coding agents. Now, we are expanding beyond software engineering. Recent progress in mathematics demonstrates what happens when models are trained in domains with massive amounts of public data and easy verifiability, making it easy to find weaknesses, generate training tasks, and iterate quickly. We build infrastructure that enables this feedback loop in other domains: we are excited about a world in which AI systems design targeted drugs, accelerate chip design, and rewrite legacy software that still runs hospitals, power grids, governments, and other critical infrastructure. Since starting, we have assembled a small team of researchers and engineers and built world-class infrastructure for post-training and synthetic data research to overcome the challenges of scaling manual data curation. Our team comes from Cursor, Google DeepMind, Meta Superintelligence, Prime Intellect, Citadel, and Jane Street. More than half are former founders. We’re fortunate to be backed by investors who share that vision, including @generalcatalyst who led our $15M Seed Round at a $300M valuation, as well as @chemistry, @svangel, @dvlamoen and individuals like @LiamFedus, @kevinweil, and @bernhardsson

Today, Justus and I are excited to introduce Proximal. We’re building the infrastructure that enables AI to improve from real-world experience. In the past year, we have grown to more than $200M in annualized revenue helping frontier labs and enterprises improve coding agents. Now, we are expanding beyond software. One of the things that I’ve enjoyed the most over the last year is working with this amazing group of people. It's an incredible feeling working alongside people who are better than you and that encourage you to do your best work. I am impressed daily at both the raw technical talent of the team but also the humility, high agency, and high trust culture we have built. Many of the team lived in the office for months over the last year which made it feel less like work but more like working on an exciting project with your best friends. It feels like (and we've been told) that it reminds them of going back to college where working with your friends on cool ideas and hacking into late nights was just so much fun. We are nowhere close to where we want to be and the ambitions we have, we’re just getting started. Thank you to everyone who has supported us thus far!

If I had to name one thing that we’ve done right at Proximal, it would be hiring. Outside of raw technical talent, I’ve never seen a place at which people care as much and have as much fun together (I’ve been told that coming to the office feels like going back to college)

Crazy that you can do these numbers without shipping “slop” RL envs… I think the future for RL is on the focused research on areas models struggle with rather than scaling one thing endlessly and @ProximalHQ is literally been build grounds up for that.. so LFG!!!

The breadth and depth of problems at Proximal made this the most exciting work I've ever been part of. Working on FrontierSWE, data gen, training, and infra with this team has been a constant reminder of what the right people can do. So much more on the horizon, stay tuned :)

Opus 5.5 is both the cheapest and highest-scoring model released by Anthropic This trend has been constant since Fable 5 - every model released after it scored higher while being cheaper than its predecessor

We partnered with @FireworksAI_HQ to bring FrontierSWE v2 to the Specialized Intelligence Index Evals are essential for improving and safely deploying AI. We're excited to support the initiative!

People with both research taste as well as good commercial instincts are true unicorns. We are hiring someone to help build our applied research function and work closely with customers. If you are an engineer or researcher that wants to learn these skills, please reach out!

GPT-6 Astra is the best-performing model on FrontierSWE Astra achieves a score of 65.5%, outperforming Fable 5.1 (56.3%) and its predecessor GPT-5.6 Sol (32.2%)

Today, we are announcing our Series B funding round, valuing the company at more than $1B. This round accelerates our next-gen Trinity models across diverse infrastructure, expands our work with the DOE and national labs on Genesis-Science-1, and enables us to build the platform teams need to build, evaluate, deploy, and operate open models in production. We are grateful to our team, partners, open-source community, and investors. Led by @Vista_Equity, Cambium Capital, and @emergencecap, with participation from AI10 Ventures, @Hitachi, IAG, @M12vc, @p7ventures, and @Wipro.

We midtrain + RL’d a trillion param LLM to analyze experimental data from our superconductor lab It’s better at it than Astra & Fable, and our scientists love it Read our first research report below

🚀 Introducing DeepSeek-V4.1-Flash: smarter, faster, more efficient. 🔹 Introducing the smallest model in our new architecture family, with native visual understanding. 🔹 Designed for greater capability, faster inference, higher throughput, and scaling to larger models. 1/6