| # | Model | Provider | Type | Score | |
|---|---|---|---|---|---|
| 1 | MAI-Thinking-1 | Microsoft | Closed | mai-thinking-1 | |
| 2 | Kimi K2.5 | Moonshot AI | Open | kimi-k2-5 | |
| 3 | Kimi K2.5 (Reasoning) | Moonshot AI | Closed | kimi-k2-5-reasoning | |
| 4 | GLM-4.7 | Z.AI · Open weight | 95.7% | ||
| 5 | MiMo-V2-Flash | Xiaomi · Open weight | 94.1% | ||
| 6 | Claude Sonnet 4.5 | Anthropic · Closed | 87% | ||
| 7 | Exaone 4.0 32B | LG AI Research · Open weight | 85.3% | ||
| 8 | GPT-5 nano | OpenAI · Closed | 85.2% | ||
| 9 | Nemotron 3 Nano Omni 30B A3B | NVIDIA · Open weight | 82.1% | ||
| 10 | LFM2.5-8B-A1B | LiquidAI · Open weight | 42.5% | ||
| 11 | MiniCPM5-1B | OpenBMB · Open weight | 40.4% |