Benchmark Almanac

MiniCPM3-4B

MiniCPM3-4B.

OpenBMB (ModelBest) Released September 3rd, 2024 Family MiniCPM The lab's model card Compare with others
5
Benchmarks with a score
0
Ranked first, out of five or more models
0
In the top three
OpenBMB (ModelBest) Sep 2024

MiniCPM3-4B

Open weights Small

No Epoch Capabilities Index score yet

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
SuperCLUE-Hard 13.7 #12 of 13 Official leaderboard

Agents and tools

BenchmarkScoreRankSource
Berkeley Function Calling Leaderboard (BFCL) 50.3% #50 of 67 Official leaderboard
BFCL V4 25.6% #51 of 65 Official leaderboard

Vision, audio and robotics

BenchmarkScoreRankSource
General-Bench 4.12 #21 of 41 Official leaderboard

Safety and security

BenchmarkScoreRankSource
SuperCLUE-DSPSafeBench 77.6 #2 of 4 Official leaderboard

What does it cost to run?

On the one leaderboard that publishes a price, its best run costs 3.8 times the median model's. There, another model comes close to its score for less.

Benchmark Score Price From cheapest to priciest Against the median Close to its score for less Close for less
BFCL V4 25.6% $54.05 per run 3.8× the median 3.8× Qwen3-0.6B 23.9% for $0.46

Who measured it?

Independent runs cover all 5 of its benchmarks.

  • An independent run only 5
Official leaderboards scored it on 5.

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