Model index
Model behaviour through one investment lens.
The board compares model behaviour against investment workflow tasks, not general preference. Every score and rank on this pass is synthetic and illustrative.
| Rank | Model | Provider | Version | Overall | Weights |
|---|---|---|---|---|---|
| 1 | OAIGPT-5.5 | OpenAI | gpt-5.5 | 83.9 | Closed |
| 2 | Claude Opus 4.8 | Anthropic | 4.8 | 82.8 | Closed |
| 3 | Gemini 3.5 Flash | 3.5 Flash | 79.5 | Closed | |
| 4 | OAIGPT-5.4 | OpenAI | gpt-5.4 | 78.3 | Closed |
| 5 | Claude Opus 4.7 | Anthropic | 4.7 | 76.8 | Closed |
| 6 | Gemini 3.1 Pro | 3.1 Pro | 75.2 | Closed | |
| 7 | Grok 4.20 | xAI | 4.20 | 74.0 | Closed |
| 8 | ZGLM-5.2 | Z.ai | 5.2 | 73.3 | Open |
| 9 | Qwen3.7 Max | Alibaba | 3.7 Max | 71.9 | Closed |
| 10 | Seed 2.0 Pro | ByteDance | 2.0 Pro | 69.6 | Closed |
| 11 | DeepSeek V4 Pro | DeepSeek | V4 Pro | 68.0 | Open |
| 12 | OAIGPT-5.2 Pro | OpenAI | gpt-5.2-pro | 66.8 | Closed |