LFM3-VL-2B
LFM3-VL-2B 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
65results on65benchmarks
- 65 vendor-reported
From 1 source · latest Oct 28, 2026 · How verification works
LFM3-VL-2B benchmark results
65 results on 65 benchmarks, grouped by capability. Each bar is the result as a share of the capability leader's; every result links to its source.
39.7% behind the leader3 of 6 ranked benchmarks measured
More results
Benchmarks outside the capability baskets. They are not ranked against a leader.
- MM-ToolSandbox (partial credit)42.50Oct 28, 2026
- 83.60Oct 28, 2026
- 85.20Oct 28, 2026
- 49.60Oct 28, 2026
- MotionBench (dev)64.00Oct 28, 2026
- MMVU-Val (Pass@1)52.90Oct 28, 2026MMVU (val)
Show 54 more resultsHide 54 results
- ActivityNet-QA (test)71.20Oct 28, 2026
- Agentic42.50Oct 28, 2026
- ChartQA Test92.70Oct 28, 2026ChartQA (test)
- CharXiv (descriptive, val)87.50Oct 28, 2026
- CharXiv (reasoning, val)70.20Oct 28, 2026
- 88.80Oct 28, 2026
- Document, OCR & Chart83.90Oct 28, 2026
- DocVQA-val96.70Oct 28, 2026DocVQA (val)
- DynaMath (average accuracy)79.20Oct 28, 2026
- Embodied reasoning64.40Oct 28, 2026
- 82.90Oct 28, 2026
- 48.20Oct 28, 2026
- General66.00Oct 28, 2026
- GUI understanding83.10Oct 28, 2026
- 65.60Oct 28, 2026
- HR-Bench (4K)82.30Oct 28, 2026
- HR-Bench (8K)76.20Oct 28, 2026
- 86.20Oct 28, 2026
- 41.20Oct 28, 2026
- Math & logical reasoning68.80Oct 28, 2026
- MATH-Vision (full)51.60Oct 28, 2026
- MindCube (tinybench)56.50Oct 28, 2026
- MLVU (dev, M-Avg)71.30Oct 28, 2026
- 72.80Oct 28, 2026
- MM-Vet (v1)57.40Oct 28, 2026
- MMBench86.10Oct 28, 2026MMBench (dev EN v1.1)
- 86.90Oct 28, 2026
- MMMU (val) (Pass@1)55.70Oct 28, 2026MMMU (val)
- 69.40Oct 28, 2026
- 86.30Oct 28, 2026
- OCRBench v2_en63.50Oct 28, 2026OCRBench v2 (En)
- OpenEQA (episodic memory, video)76.30Oct 28, 2026
- Perception75.60Oct 28, 2026
- Point-Bench73.80Oct 28, 2026
- 88.70Oct 28, 2026
- 79.30Oct 28, 2026
- 95.20Oct 28, 2026
- 49.80Oct 28, 2026
- 63.40Oct 28, 2026
- SAT (val + test)80.20Oct 28, 2026
- ScreenSpot-Pro (No tools)55.50Oct 28, 2026ScreenSpot-Pro
- 94.60Oct 28, 2026
- 92.20Oct 28, 2026
- 89.90Oct 28, 2026
- 78.60Oct 28, 2026
- 36.30Oct 28, 2026
- TextVQA-val87.90Oct 28, 2026TextVQA (val)
- V*Bench86.40Oct 28, 2026
- Video61.20Oct 28, 2026
- Visual grounding77.30Oct 28, 2026
- 63.10Oct 28, 2026
- We-Math (testmini, strict)68.60Oct 28, 2026
- 44.90Oct 28, 2026
- ZoomBench (full + crop average)62.60Oct 28, 2026
LFM3-VL-2B: common questions
Who makes LFM3-VL-2B?
LFM3-VL-2B is made by Liquid AI.
What is LFM3-VL-2B good at?
LFM3-VL-2B 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 LFM3-VL-2B been tested on?
We track 65 results for LFM3-VL-2B on 65 benchmarks from 1 source. The latest was recorded on Oct 28, 2026.
About this record
Where LFM3-VL-2B's numbers come from, and every name it appears under.
- Tracked since
- Oct 28, 2026
- Newest source mention
- Oct 28, 2026
Where the results come from
Verification: 65 scores · 0 independently verified · 65 vendor-reported. How these tiers are assigned
From 1 source on 1 site. Hugging Face supplies all of them. Bars are coloured by trust tier.
- huggingface.co65