Benchmark
Vals Code Migration
Coding benchmarks compare models on programming tasks. Scores are percentages.
- Models measured
- 76
- Scores recorded
- 76
- Sources
- 1
- IQ weight
- 2.05
Score distribution
- 32%
- Mean
- 32.5%
- Median
- 18.8%
- Std dev
- 6.8%
- 10th percentile
- 57.5%
- 90th percentile
- 0.314%
- Min
This benchmark has a weight of 2.05 for chat models, measured over 76 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 |
|---|---|---|
| 51 | Muse Spark Open· Vals AI | 21.2% |
| 52 | GLM 5.3 Flash· Vals AI | 20.5% |
| 53 | MiniMax M3· Vals AI | 19.9% |
| 54 | Gemini 3.1 Pro Preview· Vals AI | 17.3% |
| 55 | Kimi K3· Vals AI | 16.1% |
| 56 | GPT 5.4 Nano· Vals AI | 14.5% |
| 57 | MiMo V2.5· Vals AI | 14.2% |
| 58 | Qwen 3.8 27B· Vals AI | 14.2% |
| 59 | Inkling Small· Vals AI | 13.7% |
| 60 | Qwen3.7 Max· Vals AI | 13.1% |
| 61 | GPT-5.4 mini· Vals AI | 12.9% |
| 62 | Qwen3.7 Plus· Vals AI | 12.9% |
| 63 | Inkling FP4· Vals AI | 11.8% |
| 64 | Qwen3.6 Plus· Vals AI | 11.1% |
| 65 | MiniMax M2.7· Vals AI | 8.7% |
| 66 | Claude Haiku 4.5thinking· Vals AI | 7.6% |
| 67 | Kimi K2.5thinking· Vals AI | 7% |
| 68 | Grok 4.3· Vals AI | 6.8% |
| 69 | Gemini 3 Flash· Vals AI | 6.4% |
| 70 | Gemini 3.5 Flash Lite· Vals AI | 6.1% |
| 71 | Mistral Medium 3.5· Vals AI | 5.1% |
| 72 | Nemotron 3 Ultra· Vals AI | 4.9% |
| 73 | Gemini 3.1 Flash Lite· Vals AI | 4.6% |
| 74 | Mercury 2.5· Vals AI | 4.5% |
| 75 | Nemotron 3.5 Lightning 30B A3B· Vals AI | 3% |
| 76 | Grok 4.20· Vals AI | 0.314% |
Showing 51–76 of 76 results.