Run #950
success · fetched 2026-07-15 13:00:57 · 1.29 MB raw HTML · 28 models
Top quality: GPT-5.6 Sol (max) (63.4 pts)
Cost/task vs quality?
23 models · 7 on the Pareto frontier · combined quality vs Cost per Task
| #? | Pareto? | Model? | Released? | Cost/task? | Time/task? | $/Q? | Qual? | ΔTop? | Intel? | Code? | Agent? | Pen? | Score? |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | ✓ | GPT-5.6 Sol (max) OpenAI |
2026-07-09 | $1.04 | 4.08m | 0.0164 | 63.4 | 0.0 | 58.9 | 77.4 | 54.0 | 20.2 | 63.4 |
| 2 | Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback) Anthropic |
2026-06-09 | $2.75 | 5.15m | 0.0436 | 63.1 | -0.4 | 59.9 | 76.5 | 52.8 | 24.4 | 63.1 | |
| 3 | ✓ | GPT-5.6 Terra (max) OpenAI |
2026-07-09 | $0.554 | 2.20m | 0.0093 | 59.7 | -3.8 | 55.0 | 76.7 | 47.4 | 17.4 | 59.7 |
| 4 | Claude Opus 4.8 (Adaptive Reasoning, Max Effort) Anthropic |
2026-05-28 | $1.80 | 7.16m | 0.0304 | 59.0 | -4.4 | 55.7 | 74.3 | 47.2 | 22.5 | 59.0 | |
| 5 | GPT-5.5 (xhigh) OpenAI |
2026-04-23 | $0.992 | 3.86m | 0.0171 | 58.2 | -5.2 | 54.8 | 74.9 | 44.9 | 20.0 | 58.2 | |
| 6 | ✓ | Grok 4.5 (high) SpaceXAI |
2026-07-08 | $0.311 | 2.11m | 0.0054 | 57.3 | -6.1 | 53.8 | 72.4 | 45.7 | 14.9 | 57.3 |
| 7 | Claude Sonnet 5 (Adaptive Reasoning, Max Effort) Anthropic |
2026-06-30 | $1.53 | 9.75m | 0.0267 | 57.2 | -6.2 | 53.4 | 71.5 | 46.7 | 21.8 | 57.2 | |
| 8 | ✓ | GPT-5.6 Luna (max) OpenAI |
2026-07-09 | $0.209 | 1.42m | 0.0037 | 56.1 | -7.3 | 51.2 | 71.4 | 45.6 | 13.2 | 56.1 |
| 9 | GLM-5.2 (max) Z AI |
2026-06-16 | $0.468 | 3.94m | 0.0086 | 54.3 | -9.1 | 51.1 | 68.8 | 43.1 | 16.7 | 54.3 | |
| 10 | Muse Spark 1.1 (xhigh) Meta |
2026-07-09 | $0.261 | 2.99m | 0.0049 | 53.2 | -10.3 | 50.6 | 71.3 | 37.5 | 14.2 | 53.2 | |
| 11 | Gemini 3.5 Flash (high) |
2026-05-19 | $0.586 | 2.78m | 0.0111 | 52.6 | -10.8 | 50.2 | 70.1 | 37.4 | 17.7 | 52.6 | |
| 12 | Qwen3.7 Max Alibaba |
2026-05-19 | $0.985 | 1.62m | 0.0207 | 47.5 | -15.9 | 46.0 | 66.0 | 30.6 | 19.9 | 47.5 | |
| 13 | ✓ | DeepSeek V4 Pro (Reasoning, Max Effort) DeepSeek |
2026-04-24 | $0.0448 | 9.22m | 0.0010 | 46.7 | -16.8 | 44.3 | 59.4 | 36.4 | 6.5 | 46.7 |
| 14 | MiniMax-M3 MiniMax |
2026-06-01 | $0.125 | 4.29m | 0.0027 | 46.1 | -17.3 | 44.4 | 58.6 | 35.4 | 11.0 | 46.1 | |
| 15 | Gemini 3.1 Pro Preview |
2026-02-19 | $0.291 | 1.84m | 0.0064 | 45.6 | -17.9 | 46.5 | 68.8 | 21.4 | 14.6 | 45.6 | |
| 16 | ✓ | MiMo-V2.5-Pro Xiaomi |
2026-04-22 | $0.0314 | 5.78m | 0.0007 | 43.8 | -19.6 | 42.2 | 60.2 | 29.1 | 5.0 | 43.8 |
| 17 | ✓ | DeepSeek V4 Flash (Reasoning, Max Effort) DeepSeek |
2026-04-24 | $0.0223 | 5.92m | 0.0005 | 42.5 | -20.9 | 40.3 | 56.2 | 31.1 | 3.5 | 42.5 |
| 18 | Nemotron 3 Ultra 550B A55B (Reasoning) NVIDIA |
2026-06-04 | $0.234 | 1.90m | 0.0061 | 38.1 | -25.3 | 37.8 | 49.3 | 27.4 | 13.7 | 38.1 | |
| 19 | Grok 4.3 (high) SpaceXAI |
2026-04-30 | $0.139 | 2.22m | 0.0040 | 34.6 | -28.8 | 37.6 | 42.2 | 24.1 | 11.4 | 34.6 | |
| 20 | Qwen3.5 397B A17B (Reasoning) Alibaba |
2026-02-16 | $0.333 | 4.15m | 0.0098 | 33.9 | -29.5 | 33.7 | 48.2 | 19.8 | 15.2 | 33.9 | |
| 21 | Mistral Medium 3.5 Mistral |
2026-04-29 | $1.20 | 3.95m | 0.0374 | 31.9 | -31.5 | 29.9 | 46.9 | 19.0 | 20.8 | 31.9 | |
| 22 | Claude 4.5 Haiku (Reasoning) Anthropic |
2025-10-15 | $0.272 | 3.47m | 0.0091 | 30.0 | -33.5 | 29.6 | 43.9 | 16.4 | 14.3 | 30.0 | |
| 23 | gpt-oss-120b (high) OpenAI |
2025-08-05 | $0.0607 | 2.11m | 0.0027 | 22.5 | -40.9 | 23.8 | 30.4 | 13.2 | 7.8 | 22.5 |