Benchmark Almanac

MiniCPM-V-4.7-35B-A3B

MiniCPM-V-4.7-35B-A3B.

27
Benchmarks with a score
0
Ranked first, out of five or more models
0
In the top three
OpenBMB (ModelBest)

MiniCPM-V-4.7-35B-A3B

Open weights Mid-size

139.3 #188 of 412
Self-Reported Index, from the lab's own numbers. No Epoch Capabilities Index score yet.
  • HallusionBench 63.8% #18 /32
  • TextVQA 85% #1 /18

Where it ranks

Its best score on each benchmark, and where that ranks among every model with a score there. We rank each model once, at its best setting.

Knowledge and language

BenchmarkScoreRankSource
HallusionBench 63.8% #18 of 32 lab numbers Lab model card
CMMLU 90.2% Not ranked Lab model card
IFBench 47.7% Not ranked Lab model card
MM-IFEval 65.4% Not ranked Lab model card
MMLU-Pro 83.5% Not ranked Lab model card
SimpleVQA 54.1% Not ranked Lab model card
SuperGPQA 61.2% Not ranked Lab model card

Math and science

BenchmarkScoreRankSource
AIME 2026 76.5% Not ranked Lab model card

Coding

BenchmarkScoreRankSource
LiveCodeBench v6 56.6% Not ranked Lab model card

Agents and tools

BenchmarkScoreRankSource
BFCL V4 60.1% Not ranked Lab model card
tau2-bench 84.7% Not ranked Lab model card

Vision, audio and robotics

BenchmarkScoreRankSource
SuperCLUE-VLM 58.5 #14 of 14 Official leaderboard
TextVQA 85% #1 of 18 lab numbers Lab model card
BLINK 70.3% Not ranked Lab model card
CharXiv 75.6% Not ranked Lab model card
LongVideoBench 68.9% Not ranked Lab model card
MATH-Vision (MathVision) 74.2% Not ranked Lab model card
MathVerse 86.5% Not ranked Lab model card
MathVista 88% Not ranked Lab model card
MMMU 80.2% Not ranked Lab model card
MMMU-Pro 69.7% Not ranked Lab model card
MMSI-Bench 16% Not ranked Lab model card
MMStar 80.7% Not ranked Lab model card
MMT-Bench 75.5% Not ranked Lab model card
MotionBench 68.1% Not ranked Lab model card
OCRBench 91.2% Not ranked Lab model card
OCRBench v2 68.5% Not ranked Lab model card

Its lab's numbers put it at 139.3, number 188 of 412

The Self-Reported Index fits one score to the 23 capability results OpenBMB (ModelBest) reported for this model, on the scale of the Epoch Capabilities Index. Labs choose which benchmarks to show and how to run them, so read it as the lab's claim, not an independent measurement. The 90% interval runs from 136.9 to 141.8. How we compute it.

Who measured it?

Independent runs cover 1 of its 27 benchmarks. For the other 26, we have only OpenBMB (ModelBest)'s own number or a claim Epoch AI collected.

  • An independent run only 1
  • A claim only 26
Official leaderboards scored it on 1.

Sources: Official leaderboards, Epoch AI (CC BY 4.0) and labs’ model cards

Its launch post

The benchmarks the lab chose to headline, with the scores it gave.