Big techChina
Alibaba's Qwen model team, separate from DAMO Academy, which extended its vision-language models to robot action (Qwen-VLA) and driving (Qwen-Drive).
The Qwen Team is Alibaba's foundation-model group behind the Qwen language and vision-language model families. In May 2026 it published Qwen-VLA, which extends Qwen3.5-4B with a 1.15B-parameter action decoder for manipulation and navigation, and in August 2026 it released Qwen-Drive-1.0, an Apache-2.0 driving model with perception and planning modules.
Sources: github.com (19 Oct 2026) · arxiv.org (28 May 2026) · huggingface.co (19 Oct 2026)
Corporate division · Model team within Alibaba Group; distinct from Alibaba DAMO Academy; no separate funding.
#24 of 104 model companies · score 65.3 / 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: QwenLM/Qwen-VLA (774 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
Vision-language foundation model for autonomous driving built on Qwen3.5-4B with a BEV perception head and a Planning Expert that outputs ego trajectories.
No benchmark results on a shared public benchmark have been published for this model.
Closed weightsVLA
Unified VLA that adds a DiT flow-matching action decoder to Qwen3.5-4B and covers manipulation, navigation and trajectory prediction with embodiment-aware prompts.
| Benchmark | Result | Source |
|---|---|---|
| LIBEROLIBERO success rate (%), as reported | 97.9% | arxiv.orgAbstract and repo table, Qwen-VLA-Instruct as one generalist model; suite breakdown not checked. As of 28 May 2026 |
| SimplerEnv WidowXWidowX (Bridge) average success rate (%) | 73.7% | arxiv.orgAbstract and repo table, Qwen-VLA-Instruct. As of 28 May 2026 |
