Robot makerShenzhen, ChinaFounded 2023
Builds the Quanta wheeled bimanual and humanoid robots and trains the WALL model family, including the open-weight WALL-OSS series.
X Square Robot is a Shenzhen-based embodied AI company founded in December 2023 by CEO Wang Qian. It develops robot foundation models, including WALL-B (introduced April 2026) and the open-sourced WALL-OSS and WALL-WM, alongside robot hardware. In June 2026 it closed four consecutive rounds ending in a Series C, reaching a valuation above $2.8 billion.
Sources: therobotreport.com (30 Jun 2026) · caixinglobal.com (26 Feb 2026)
Private company · $140M disclosed in USD across 1 of 4 listed rounds
| Date | Round | Amount | Lead | Source |
|---|---|---|---|---|
| Sep 2025 | Series A+ | Nearly RMB 1B | Alibaba Cloud | caixinglobal.com, 26 Feb 2026 |
| Jan 2026 | Series A++ | About RMB 1B (about $140M) | n/a | frontier-enterprise.com, 22 Jan 2026 |
| Feb 2026 | Undisclosed | Several hundred million RMB | SAIC Capital; CICC-SAIC Motor jointly managed fund | caixinglobal.com, 26 Feb 2026 |
| Jun 2026 | Series CValuation: Over $2.8B | Undisclosed | n/a | therobotreport.com, 30 Jun 2026 |
#42 of 104 model companies · score 39.7 / 100 · updated 19 Oct 2026
Each bar shows an input from 0 to 100% before weighting. 2 of 4 listed models have open weights. Most-starred official repository: X-Square-Robot/wall-x (1.3k 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 weightsVLA
4B VLA on a 3B VLM backbone pretrained on more than 20 embodiments; reports 60.5% average task progress after fine-tuning on 15 real-robot tasks.
No benchmark results on a shared public benchmark have been published for this model.
Open weightsVLA
End-to-end embodied foundation model on a Qwen2.5-VL-3B backbone that unifies instruction reasoning, subgoal decomposition and action generation; released in FLOW and FAST variants.
No benchmark results on a shared public benchmark have been published for this model.
Weights: n/aWorld model
World action model that uses semantically coherent action events, rather than fixed-length chunks, as the unit for joint future-video and action prediction.
No benchmark results on a shared public benchmark have been published for this model.
Closed weightsVLA
Proprietary model described as combining vision, language, action and physical prediction in a single network trained jointly from scratch; successor to WALL-A.
No benchmark results on a shared public benchmark have been published for this model.
