BLXBench Cost category
Code · 2026-05-10
Cost category component of the BLXBench v2 public leaderboard. It scores cost-aware correctness and efficient API spend per successful task; higher percentage scores indicate better category performance.
Top models (higher is better)
| Model | Score |
|---|---|
| Opus 4.8 | 92.7 |
| GPT-5.3-Codex | 92.7 |
| Grok Build 0.1 | 92.7 |
| Nemotron 3 Super 120B A12B | 92.7 |
| Opus 4.7 | 91.3 |
| GPT-5.5 | 91.3 |
| Granite 4.1 8B | 90.7 |
| Grok 4.3 | 90.7 |
| MiMo-V2.5 | 90.7 |
| Mistral Small 4 | 90.7 |
| CoBuddy | 90.0 |
| DeepSeek-V4-Pro | 90.0 |
| Mistral Medium 3.5 | 90.0 |
| Qwen3.6-Flash-2026-04-16 | 90.0 |
| Qwen3.7-Plus | 90.0 |
| Nemotron 3 Nano 30B A3B | 89.3 |
| Qwen3.7-Max | 89.3 |
| GLM-5.1 | 88.7 |
| Claude Fable 5 | 88.0 |
| Gemini 3.1 Flash-Lite | 88.0 |
| Kimi K2.6 | 88.0 |
| MiniMax M3 | 88.0 |
| North Mini Code | 88.0 |
| DeepSeek-V4-Flash | 87.3 |
| MiMo-V2.5-Pro | 87.3 |
| Nemotron 3 Nano Omni 30B A3B | 85.3 |
| Ring 2.6 1T | 84.8 |
| GLM-5.2 | 84.7 |
| Kimi K2.7 Code | 82.7 |
| MiniMax M2.7 | 78.0 |
| Step 3.7 Flash | 66.0 |
| Gemini 3.5 Flash | 56.7 |