Research labNew York, NY, United States
Lerrel Pinto's General Robotics & AI Lab at New York University; released the BAKU multi-task policy.
GRAIL (General Robotics & AI Lab) at New York University is run by Lerrel Pinto and focuses on robot learning, representation learning and open-source robots. In June 2024 it released BAKU, a transformer architecture for multi-task policy learning evaluated on LIBERO-90, Meta-World, DM Control and 30 tasks on a real xArm. Pinto's site states he also works at Meta Superintelligence Labs after Meta acquired his company Assured Robot Intelligence in May 2026.
Sources: lerrelpinto.com (19 Oct 2026) · arxiv.org (11 Jun 2024)
Research lab · University research lab at NYU; no separate funding.
#57 of 104 model companies · score 28.8 / 100 · updated 19 Oct 2026
Each bar shows an input from 0 to 100% before weighting. 1 of 2 listed models have open weights. Most-starred official repository: haritheja-e/robot-utility-models (248 stars, 19 Oct 2026).
The score (0–100) measures public, checkable evidence about a company's robot models. It does not measure private capability. A company that publishes no benchmark results and no weights can still have strong models. Every input is computed from the data on each company page, and every number there links to its source.
| Input | Weight | How it is measured |
|---|---|---|
| Published benchmarks | 35% | For each scored benchmark, the company's best result divided by the best result any company has published, averaged over the benchmarks it reports. No published results scores 0. Results come only from the model's own paper, model card or the official benchmark leaderboard. Scored: LIBERO (4-suite average success, %); SimplerEnv Google Robot, visual matching (%); SimplerEnv Google Robot, variant aggregation (%); SimplerEnv WidowX / Bridge (%); CALVIN ABC→D (average length, 0–5); RoboArena (pairwise rating). |
| Open weights | 20% | Share of the company's listed models whose weights the public can download. |
| Adoption | 20% | GitHub stars on the company's most-starred official model repository, on a log scale relative to the most-starred repository in the list. Hugging Face downloads are shown on company pages but not scored, because gated repositories report zero. |
| Released models | 15% | Number of publicly announced, named models, capped at 5. |
| Disclosed funding | 10% | The sum of the funding rounds listed on the company's profile that were disclosed in USD, on a log scale from $10M to the largest sum in the list. Rounds reported only in other currencies are listed but not converted or counted. This input applies only to private startups and robot makers. For big-tech divisions, research labs and listed companies it is n/a, and its weight is shared across the other inputs in proportion. |
Sources. arXiv papers and official model cards for benchmarks and parameters; official leaderboards (RoboArena); the GitHub and Hugging Face APIs for stars and downloads; company announcements and established press for funding. Missing values are shown as n/a, never estimated.
Updated 19 Oct 2026. Benchmark protocols differ between papers, for example in fine-tuning and number of trials, so small gaps between close scores are not meaningful.
Open weightsPolicy
Set of five task-specific imitation-learned policies that deploy zero-shot in unseen homes on the Hello Robot Stretch, with an mLLM verifier for retrying.
No benchmark results on a shared public benchmark have been published for this model.
Weights: n/aPolicy
Transformer architecture for multi-task robot policies that combines observation trunks, FiLM language conditioning and action chunking; reported model sizes range from about 4M to 114M parameters.
No benchmark results on a shared public benchmark have been published for this model.
