NewsWe've raised $19M to simulate the world's work

Experience,
Beyond Imitation.

We build worlds for models to gain experience in.

Each of our worlds is a working simulation of the tools, software and systems a job runs on, built around the exact tasks and workflows people do in it. Models train inside them on that same work, so the skills they learn transfer directly to real-world deployment.

Benchmarks

How frontier models do in our worlds

All research →
Finance & Accounting: 75 worldsEngineering & Design: 64 worldsLegal & Compliance: 42 worldsHealthcare Administration: 28 worldsMarketing: 23 worldsInsurance / Claims: 23 worldsSoftware & IT: 19 worldsSales & Revenue: 12 worldsOperations & Management: 7 worldsHR & People Ops: 7 worlds

Computer Work Val v1.0

300 worlds from the real work of people earning over $100k a year. Average pass@1 at high reasoning effort.

modelavg pass@1
Gemini 4 Argon
[high]
58.5%
Claude Opus 5.5
[high]
52.5%
GPT-6 Astra
[high]
43.1%
Grok 4.7
[high]
39.9%
Muse Wonder
[high]
39.6%
Full results →

Web Clone Bench

Rebuild a working web app by studying a running copy, on 9 reference apps. Full score at each model's best effort.

modelfull
Claude Fable 5.2
[low]
62.0%
Claude Opus 5.5
[xhigh]
34.2%
GPT-6 Astra
[max]
17.5%
Claude Fable 5.1
[high]
13.3%
GPT-6.1 Sol
[max]
12.6%
Full results →

Careers

Work with us

We're hiring across infrastructure, research and engineering in San Francisco.

All open roles →

Partner with Plato

Get access to our worlds.

Train and evaluate your models across the worlds we've built and the work inside them. Reach us at partner@plato.so or tell us about your team.