Kimi VL A3B Instruct
Kimi VL A3B Instruct is behind the leaders in multimodal tasks. Too few results yet to rate reasoning, coding, agentic tasks, safety, long context, math, multilingual tasks, instruction following, or factuality.
Capability profile
Bars show the model's median result as a share of the leading model's, per capability. Select a row to see its results.
Too few results yet to rate Reasoning, Coding, Agentic, Safety, Long Context, Math, Multilingual, Instruction Following or Factuality.
Price
No current price is tracked for this model. See the rate card
Evidence
38results on30benchmarks
- 38 vendor-reported
From 2 sources · latest Jun 12, 2026 · How verification works
Kimi VL A3B Instruct benchmark results
38 results on 30 benchmarks, grouped by capability. Each bar is the result as a share of the capability leader's; every result links to its source.
50.7% behind the leader3 of 6 ranked benchmarks measured
- MathVista68.70Jun 12, 2026MathVista (Pass@1)
- MMMU57.00Jun 5, 2026MMMU (val, Pass@1)
- MMMU-Pro36.00Jun 5, 2026MMMU-Pro (Pass@1)
Show 1 more multimodal resultHide 1 multimodal result
- MathVista68.60Jun 5, 2026MathVista_MINI (Pass@1)
0 of 7 ranked benchmarks measured
- OSWorld-Verified8.22Jun 12, 2026OSWorld (Pass@1)
More results
Benchmarks outside the capability baskets. They are not ranked against a leader.
- 61.70Jun 12, 2026
- 864.00Jun 12, 2026
- 68.10Jun 12, 2026
- 82.90Jun 12, 2026
- 31.70Jun 12, 2026
- 37.40Jun 12, 2026
Show 27 more resultsHide 27 results
- 84.90Jun 12, 2026
- 57.30Jun 12, 2026
- 78.50Jun 12, 2026
- 83.20Jun 12, 2026
- 64.50Jun 12, 2026
- MATH-Vision21.40Jun 12, 2026MathVision (Pass@1)
- MATH-Vision21.70Jun 5, 2026MATH-Vision (Pass@1)
- 74.20Jun 12, 2026
- 83.10Jun 12, 2026
- 35.10Jun 12, 2026
- MMMU (val) (Pass@1)57.00Jun 12, 2026MMMU-Val (Pass@1)
- 61.30Jun 12, 2026
- 66.70Jun 5, 2026
- 66.70Jun 12, 2026
- MMVU52.70Jun 5, 2026MMVU (Pass@1)
- 52.20Jun 12, 2026
- 867.00Jun 12, 2026
- 41.60Jun 5, 2026
- 34.50Jun 12, 2026
- 35.40Jun 5, 2026
- ScreenSpot-V292.80Jun 12, 2026ScreenSpot-V2 (Acc)
- 67.80Jun 12, 2026
- 72.60Jun 12, 2026
- VideoMME (w sub.)72.70Jun 5, 2026Video-MME (w/ sub.)
- VideoMMMU52.60Jun 12, 2026VideoMMMU (Pass@1)
- VideoMMMU52.10Jun 5, 2026VideoMMMU (Pass@1)
- 10.40Jun 12, 2026
Kimi VL A3B Instruct: common questions
Who makes Kimi VL A3B Instruct?
Kimi VL A3B Instruct is made by Moonshot.
When was Kimi VL A3B Instruct released?
Kimi VL A3B Instruct's weights were first published on Hugging Face on Apr 9, 2025.
What is Kimi VL A3B Instruct good at?
Kimi VL A3B Instruct is behind the leaders in multimodal tasks. Too few results yet to rate reasoning, coding, agentic tasks, safety, long context, math, multilingual tasks, instruction following, or factuality.
How many benchmarks has Kimi VL A3B Instruct been tested on?
We track 38 results for Kimi VL A3B Instruct on 30 benchmarks from 2 sources. The latest was recorded on Jun 12, 2026.
About this record
Where Kimi VL A3B Instruct's numbers come from, and every name it appears under.
- Tracked since
- Jun 5, 2026
- Newest source mention
- Jun 5, 2026
Where the results come from
Verification: 38 scores · 0 independently verified · 38 vendor-reported. How these tiers are assigned
From 2 sources on 1 site. Hugging Face supplies all of them. Bars are coloured by trust tier.
- huggingface.co38