Stability AI: stable-code-3b
1 benchmarks
Leaderboard position
#750
Overall rank (of 1121 models)
566
Elo rating
1
Benchmarks
Top 66.9%
Percentile
In Stability AI (of 6 models)
Above: flock-io: this-that-model-1.2 #749
Below: ElevenLabs: multilingual-v2 #751
Benchmark highlights
Best benchmark
-(CoderPPL)
-2.1903
Worst benchmark
-(CoderPPL)
-2.1903
Nearest competitor from another company
The model from another company with the most similar benchmark scores.
appeared as nearest in 1 of 1 benchmarks
Benchmarks
Predicted scores
Estimated from 25 similar models and the global Elo rating. These are projections, not measured results.
| Benchmark | Predicted | Confidence | Based on |
|---|---|---|---|
| MATH-500 | ~81.38 | 50% | — |
| AIME 2025 | ~56.64 | 50% | — |
| AIME 2026 | ~67.92 | 50% | — |
| HMMT Feb 2026 | ~59.41 | 50% | — |
| GPQA Diamond | ~51.74 | 50% | — |
| MMLU-Pro | ~54.15 | 50% | — |
| Humanity's Last Exam (no tools) | ~6.5 | 50% | — |
| IFEval | ~78.33 | 50% | — |
| IFBench | ~48.56 | 50% | — |
| HumanEval | ~65.33 | 50% | — |
| MBPP | ~66.07 | 50% | — |
| Terminal-Bench v2.1 | ~11.51 | 50% | — |
| MMLU | ~60.24 | 50% | — |
| SWE-bench Verified | ~30.14 | 50% | — |
| SWE-bench Pro | ~9.7 | 50% | — |
| SciCode | ~22.42 | 50% | — |
| AA-LCR | ~37.46 | 50% | — |
| Artificial Analysis Intelligence Index v5.0.1 | ~12 | 50% | — |
| Humanity's Last Exam (with tools) | ~11.9 | 50% | — |
| GSM8K | ~71.97 | 50% | — |
| LiveCodeBench v6 | ~49.34 | 50% | — |
| MMMU-Pro | ~47.54 | 50% | — |
| MATH | ~48.66 | 50% | — |
| BBH | ~58.93 | 50% | — |
| ARC-Easy | ~75.97 | 50% | — |