Benchmark
Vals Code Migration v1.1
Coding benchmarks compare models on programming tasks. Scores are percentages.
- Models measured
- 29
- Scores recorded
- 29
- Sources
- 1
- IQ weight
- 1.96
Score distribution
- 49.9%
- Mean
- 51.6%
- Median
- 22%
- Std dev
- 20.9%
- 10th percentile
- 80.6%
- 90th percentile
- 8.8%
- Min
This benchmark has a weight of 1.96 for chat models, measured over 29 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.
Leaderboard
| # | Model | Score |
|---|---|---|
| 1 | Claude Sonnet 5.5· Vals AI | 86.9% |
| 2 | Claude Fable 5.2· Vals AI | 83.3% |
| 3 | Claude Opus 5.5· Vals AI | 81% |
| 4 | GPT-6 Astra· Vals AI | 80.6% |
| 5 | GPT-6.1 Sol· Vals AI | 77.3% |
| 6 | Muse Wonder· Vals AI | 72.4% |
| 7 | Gemini 4 Argon· Vals AI | 70.8% |
| 8 | DeepSeek V4.1 Flash 1017· Vals AI | 67.5% |
| 9 | Muse Wonder Full· Vals AI | 62.8% |
| 10 | DeepSeek V4.1 Flash· Vals AI | 58.9% |
| 11 | GLM 5.3· Vals AI | 55.1% |
| 12 | Grok 4.7· Vals AI | 53.2% |
| 13 | Claude Haiku 5.5· Vals AI | 52.4% |
| 14 | DeepSeek V4 Pro 0813· Vals AI | 52% |
| 15 | GLM 5.3 Flash· Vals AI | 51.6% |
| 16 | MiMo V2.6 Pro· Vals AI | 48.8% |
| 17 | Step 5 Preview· Vals AI | 47% |
| 18 | GPT-6 Luna· Vals AI | 46.7% |
| 19 | Hy4 preview· Vals AI | 45.4% |
| 20 | MiMo V2.6 Flash· Vals AI | 39.6% |
| 21 | Mercury 3· Vals AI | 39.3% |
| 22 | Gemini 3.8 Flash· Vals AI | 38.6% |
| 23 | Kimi K3· Vals AI | 28.6% |
| 24 | Muse Spark Open· Vals AI | 26.9% |
| 25 | Beam 1.0 501B A23B· Vals AI | 22.8% |
| 26 | MiniMax M3· Vals AI | 20.9% |
| 27 | NVIDIA Nemotron 3.5 Super VL 120B· Vals AI | 16.4% |
| 28 | Inkling Small· Vals AI | 10.4% |
| 29 | Inkling FP4· Vals AI | 8.8% |