Osmosis
A reinforcement-learning post-training platform.
A reinforcement-learning post-training platform. You train a smaller model to do one task well, with Osmosis staff alongside you, and the homepage says it can take your evaluation results and retrain the model automatically, "as often as every hour". There is an MIT-licensed Python SDK and command-line tool for building agent rollouts, running evaluations and submitting training runs.
NO LINKS READ YET · no published dossier names Osmosis beside another company
Company details
| Industry | AI & Machine Learning |
| Funding | |
| Stage | Seed |
| Website | Visit Osmosis |
Others in the same industry
5 namesWhere Osmosis sits against the other names we cover on this beat. Each line is that company’s verdict, not a summary of it.
Amazon (AWS)
AMZNThe 2026-07-30 print settled four of the five questions the prior dossier said it could
Cash $78.2B
Meta AI / FAIR
METAThe bear case arrived a year early and the bull case grew a new leg in the same quarter
Cash $90.3B
Meta Platforms
METAThe buildout stopped being paid for by the ad business and started being paid for by the capital markets
Cash $90.3B
Alphabet
GOOGLThe market spent two years pricing Alphabet as the AI loser; FY2025 (+15% op income, Cloud margin to 24%, Gemini at 750M users, the antitrust gun most…
Cash $55.9B
Palantir
PLTRA genuinely great business priced as a perfect one
Cash $2.0B
Comments and ideas
An idea we take on becomes an entry in the work log, with your name on it.