Benchmarks · math
Riemann-Bench
Uses Surge AI's first-party extreme-math leaderboard score. Score: score. Collected by Jake Boggs' benchmark dashboard.
41 models · 43 results · data as of 2026-10-29 · surgehq.ai ↗
Leaderboard
Gemini 4 Carbon high
All results
43 rows| # | Model | Effort | Riemann-Bench |
|---|---|---|---|
| 1 | Gemini 4 Carbon | high | 80.0%launch |
| 2 | GPT-5.6 Sol | max | 74.4% |
| 3 | GPT-6 Astra | max | 72.0% |
| 4 | Opus 5.5 | max | 69.6% |
| 5 | Opus 5 | max | 68.0% |
| 6 | GPT-5.6 Terra | max | 66.4% |
| 7 | Fable 5.1 | max | 65.6% |
| 8 | GPT-6 Sol | max | 64.8% |
| 9 | Fable 5 | max | 60.0% |
| 10 | GPT-6 Luna | max | 59.2% |
| 11 | GPT-5.5 | xhigh | 55.2% |
| 12 | Gemini 3.8 Flash | high | 51.2% |
| 13 | Opus 4.8 | max | 47.2% |
| 14 | GPT-5.4 | xhigh | 41.6% |
| 15 | Gemini 3.7 Flash | high | 39.2% |
| 16 | DeepSeek V4 Pro | max | 38.4% |
| 17 | Grok 4.5 | high | 38.4% |
| 18 | GPT-5.2 | xhigh | 37.6% |
| 19 | Kimi K3 | max | 37.6% |
| 20 | Gemini 3.5 Flash | high | 36.8% |
| 21 | Gemini 3.1 Pro | high | 33.6% |
| 22 | Opus 4.7 | max | 32.8% |
| 23 | Gemini 3.6 Flash | high | 30.4% |
| 24 | Muse Spark 1.3 | max | 28.0% |
| 25 | Opus 4.6 | max | 27.2% |
| 26 | DeepSeek V4.1 Flash | max | 24.0% |
| 27 | Hy4 | high | 24.0% |
| 28 | DeepSeek V4 Flash | high | 23.2% |
| 29 | Muse Spark 1.2 | max | 23.2% |
| 30 | Muse Spark 1.1 | max | 20.8% |