Expert data and evals for frontier labs. Benchmarks. Enterprise AI on your expert judgment. From the team at @joinHandshake.
San Francisco, CA
Joined September 2026
- Tweets20
- Following50
- Followers278
- Likes36
Pinned Tweet
We’re Handshake AI 👋🏼
We work with frontier labs on the expert data, environments and evaluations their models train on. We publish the benchmarks. And we bring the same approach to enterprises, so they can build AI on their own experts’ judgment.
Post-training Qwen3.8 27B on Handshake AI’s life science data raised its accuracy on the VIALS benchmark from 20.5% to 27.1%. Better than GPT-6 Luna High and 6.2pp higher than Kimi K3 Medium at 50 times lower cost. Thank you to @Osmosis_AI for the collab.
Handshake AI’s data and RL environments help research teams bridge the gap to the frontier.
Post-training Qwen3.8 27B on @joinHandshakeAI life science data raised accuracy on life science tasks by 7% (~20% -> 27%) on ATLAS-VIALS bringing it to the Pareto frontier of OSS models at 1/50th the cost of Kimi K3 and with similar accuracy as GPT-6 Luna.
The gains transfer to other benchmarks, e.g.
LAB-Bench FigQA: 75.3%→77.2%
MicroVQA: 63.8%→68.3%
PerceptionBench: 53.8%→54.8%
HLE multimodal: 17.5%→19.3%
We saw similar gains on Nemo where our data doubled it's accuracy on VIALS. Full blog post here: joinhandshake.com/research/a…
See more on ATLAS VIALS on @ArtificialAnlys's page here: artificialanalysis.ai/evalua…
Artificial Analysis ran our VIALS benchmark as an agent eval, VIALS-AA.
Give a model tools, and scores nearly double, from about 25% to 47%.
Even then, a scientist still finishes these tasks faster and more accurately.
Take away its tools, and Claude Fable 5.2 scores below Claude Haiku 5.5 with tools on VIALS-AA; across models tool use nearly doubles scores, from ~25% to ~47% on average
VIALS-AA is our implementation of @joinHandshakeAI's VIALS benchmark for visual reasoning in life sciences research. By default, models work as agents: they think, run code and view images in a sandbox. We reran VIALS-AA in a one-shot format: one request with the question and images, and no tools.
Lower cost models in an agent harness can beat single-shot large models on both cost and performance in visual reasoning and interpretation of scientific artifacts.
Claude Fable 5.2 (max) leads VIALS-AA at 67% with tools, at a cost of $4.20 per task; without tools it scores 44% at $0.73 per task. Claude Haiku 5.5 (max) scores 49% with tools, at $0.07 per task: higher than Fable without tools, at about a tenth of the cost.
2
11
508
Handshake AI retweeted
Our VIALS benchmark reveals clear weaknesses of current vision encoders. Human scientists can accurately do these tasks in minutes without using tools.
While it's good that agents with the right tools can perform significantly better in these VQA tasks, I believe this indicates we have a long way to go in computer vision research. For instance, I think Astra represented a fundamental breakthrough in computer vision when it was released (only VLM to reach 50% on VIALS), and we need more fundamental breakthroughs.
Generalization in image space seems to be harder than in text space, thanks to the fact that society has compressed the world into language so effectively.
Take away its tools, and Claude Fable 5.2 scores below Claude Haiku 5.5 with tools on VIALS-AA; across models tool use nearly doubles scores, from ~25% to ~47% on average
VIALS-AA is our implementation of @joinHandshakeAI's VIALS benchmark for visual reasoning in life sciences research. By default, models work as agents: they think, run code and view images in a sandbox. We reran VIALS-AA in a one-shot format: one request with the question and images, and no tools.
Lower cost models in an agent harness can beat single-shot large models on both cost and performance in visual reasoning and interpretation of scientific artifacts.
Claude Fable 5.2 (max) leads VIALS-AA at 67% with tools, at a cost of $4.20 per task; without tools it scores 44% at $0.73 per task. Claude Haiku 5.5 (max) scores 49% with tools, at $0.07 per task: higher than Fable without tools, at about a tenth of the cost.
Handshake AI retweeted
So proud of our team! HandshakeAI didn't exist 2 years ago.
Today we’re on the inaugural AI 100 from @NewcomerMedia and @BessemerVP.
We’re honored to be part of the inaugural 2026 AI 100, @NewcomerMedia and @BessemerVP's ranking of the top 100 private AI companies in the world.
The credit goes to the team and the experts in our network who do the work.
We’re honored to be part of the inaugural 2026 AI 100, @NewcomerMedia and @BessemerVP's ranking of the top 100 private AI companies in the world.
The credit goes to the team and the experts in our network who do the work.
Full list: ai100.newcomer.co
Models are confident narrators, which can be a good thing.
When does it become a problem?
When they’re confidently wrong.
Tune in to @jomulr and @NickMagazine1 👇🏽
In this example, @NickMagazine1 goes through Opus 5.5 analyzing an image from a biopsy. In this case, the model chose the wrong answer, and that wrong answer would cause harm to a real patient.
It’s a great demonstration of why using AI models in medicine can be so high-stakes.
Handshake AI retweeted
Handshake AI now has its own home on X!
Follow for our research and benchmarks, our work with frontier labs, and the enterprise AI we’re building on top of it. 👇
@joinHandshakeAI
Handshake AI retweeted
What does it take to push frontier models forward?
@latentspacepod's first deep dive with an AI data company. Our team walked through our own datasets, life sciences, finance and safety, step by step.
Worth your time if you care about progress and alignment of AI.
🆕 Our first deep dive with an AI data company!
The frontier data space is exploding. @joinHandshake has become one of the leading partners for frontier labs, going from launch to $1B in GMV in ~a year, with @sacrainc estimating net revenue AFTER contractor payouts at about $450M per year.
youtu.be/v5IJwVnTxzg
What are frontier labs buying at that scale? Who is building the RL environments that build the models that you build with? How hard are evals getting?
For the first time in our pod, we're speaking across a company, from cofounders @GarrettLord and @Sbhaiwala03 to leaders of Handshake Labs, Life Sciences, Finance, and Alignment. To cover this required assembling our complete team of Science, Engineering, Research and Finance hosts.
For the first time, we walked through the tasks live:
🌸 50:59 · Nick puts nine brain-tissue slides on screen. Can you pick the two with the most advanced vessel necrosis?
🟢 1:25:45 · Collin traces a bad valuation back to one line of Python. The model knows the tax rule when asked. Can it recognize where to apply it?
🟠 2:07:21 · Steve and Nishant follow an agent through the files it read, the copies it checked, and the shell history it cleared.
latent.space/handshake
