ObviousBench
GPQA.ai
The dedicated results view for one combined ranking across every fully accepted mean@N protocol. Filters preserve the official rank.
Results 0.1.14Data through Oct 8, 2026
Primary results
Official competition rank across every accepted protocol. Filters narrow the view; they never recompute rank.
Cost available for 252 of 268 configurations · 16 unavailable · prices dated 2026-10-10. USD for the full 189-question Clean run, one attempt each. Explore the Cost–score frontierDownload cost data
| Rank Rank movement Compares this configuration's standard competition rank on Clean 189 with its paired rank on Original 198.
Ranks use exact scores. Tied ranks list lower Cost first, with unavailable Cost last. Filters never recompute them. | Model | Clean score | Cost | Clean change | Benchmark source Benchmark source classes How each score entered the combined ranking.
|
|---|---|---|---|---|---|
| 1— |
GPT 6 AstraOpenAI · mean@1 · maxNative transport composite
|
Clean score98.9% | Cost$16.72 | Clean change+3.0 ppfrom 96.0% | Benchmark sourceNative transport composite |
Evidence detailExpand this record to load its answer-safe evidence detail. | |||||
| 2↑2 |
GPT 6.1 SolOpenAI · mean@1 · maxNative transport composite
|
Clean score98.4% | Cost$3.58 | Clean change+3.5 ppfrom 94.9% | Benchmark sourceNative transport composite |
Evidence detailExpand this record to load its answer-safe evidence detail. | |||||
| 2— |
GPT 6 SolOpenAI · mean@1 · maxNative transport composite
|
Clean score98.4% | Cost$4.61 | Clean change+3.0 ppfrom 95.5% | Benchmark sourceNative transport composite |
Evidence detailExpand this record to load its answer-safe evidence detail. | |||||
| 2— |
GPT 6 Astra ProOpenAI · mean@1 · maxComplete native primary
|
Clean score98.4% | Cost$21.81 | Clean change+3.0 ppfrom 95.5% | Benchmark sourceComplete native primary |
Evidence detailExpand this record to load its answer-safe evidence detail. | |||||
| 5↑2 |
Gemini 3.7 FlashGoogle · mean@1 · highNative transport composite
|
Clean score97.4% | Cost$2.02 | Clean change+2.9 ppfrom 94.4% | Benchmark sourceNative transport composite |
Evidence detailExpand this record to load its answer-safe evidence detail. | |||||
| 5↓1 |
Gemini 3.8 FlashGoogle · mean@1 · highNative transport composite
|
Clean score97.4% | Cost$4.33 | Clean change+2.4 ppfrom 94.9% | Benchmark sourceNative transport composite |
Evidence detailExpand this record to load its answer-safe evidence detail. | |||||
| 5↑5 |
Gemini 3.6 FlashGoogle · mean@1 · highNative transport composite
|
Clean score97.4% | Cost$6.10 | Clean change+3.4 ppfrom 93.9% | Benchmark sourceNative transport composite |
Evidence detailExpand this record to load its answer-safe evidence detail. | |||||
| 5↑5 |
GPT 5.5 Pro pre-releaseOpenAI · mean@7 · xhighDerived uniform
|
Clean score97.4% | Cost$196.29 | Clean change+3.4 ppfrom 93.9% | Benchmark sourceDerived uniform |
Evidence detailExpand this record to load its answer-safe evidence detail. | |||||
| 9— |
GPT 5.5 pre-releaseOpenAI · mean@8 · xhighStrict
|
Clean score97.2% | Cost$31.35 | Clean change+3.2 ppfrom 94.0% | Benchmark sourceStrict |
Evidence detailExpand this record to load its answer-safe evidence detail. | |||||
| 10↓4 |
Gemini 3.1 Pro PreviewGoogle · mean@8 · defaultTargeted recovery
|
Clean score97.0% | Cost$14.14 | Clean change+2.5 ppfrom 94.5% | Benchmark sourceTargeted recovery |
Evidence detailExpand this record to load its answer-safe evidence detail. | |||||
| 11↑1 |
GPT 5.6 SolOpenAI · mean@1 · maxComplete native primary
|
Clean score96.8% | Cost$7.44 | Clean change+3.4 ppfrom 93.4% | Benchmark sourceComplete native primary |
Evidence detailExpand this record to load its answer-safe evidence detail. | |||||
| 12↑1 |
GPT 5.4OpenAI · mean@8 · xhighTargeted recovery
|
Clean score96.7% | Cost$15.98 | Clean change+3.4 ppfrom 93.2% | Benchmark sourceTargeted recovery |
Evidence detailExpand this record to load its answer-safe evidence detail. | |||||
| 13↑7 |
Qwen3.8 Max 0902Qwen · mean@1 · xhighComplete native primary
|
Clean score96.3% | Cost$13.58 | Clean change+3.9 ppfrom 92.4% | Benchmark sourceComplete native primary |
Evidence detailExpand this record to load its answer-safe evidence detail. | |||||
| 13↓6 |
Gemini 3.1 Pro PreviewGoogle · mean@1 · highStrict
|
Clean score96.3% | Cost$14.97 | Clean change+1.9 ppfrom 94.4% | Benchmark sourceStrict |
Evidence detailExpand this record to load its answer-safe evidence detail. | |||||
| 13↑1 |
Grok 4.7xAI · mean@1 · xhighNative transport composite
|
Clean score96.3% | Cost$22.94 | Clean change+3.4 ppfrom 92.9% | Benchmark sourceNative transport composite |
Evidence detailExpand this record to load its answer-safe evidence detail. | |||||
| 16↑2 |
Gemini 3.5 FlashGoogle · mean@8 · highStrict
|
Clean score96.1% | Cost$9.76 | Clean change+3.3 ppfrom 92.8% | Benchmark sourceStrict |
Evidence detailExpand this record to load its answer-safe evidence detail. | |||||
| 17↓3 |
DeepSeek V4 FlashDeepSeek · mean@1 · maxComplete native primary
|
Clean score95.8% | Cost$2.06–$4.11 | Clean change+2.8 ppfrom 92.9% | Benchmark sourceComplete native primary |
Evidence detailExpand this record to load its answer-safe evidence detail. | |||||
| 17↓3 |
Claude Sonnet 5.5Anthropic · mean@1 · highDirect API run
|
Clean score95.8% | Cost$2.10 | Clean change+2.8 ppfrom 92.9% | Benchmark sourceDirect API run |
Evidence detailExpand this record to load its answer-safe evidence detail. | |||||
| 17↑3 |
Grok 4.6xAI · mean@1 · highComplete native primary
|
Clean score95.8% | Cost$6.51 | Clean change+3.3 ppfrom 92.4% | Benchmark sourceComplete native primary |
Evidence detailExpand this record to load its answer-safe evidence detail. | |||||
| 20↓1 |
Gemini 3 Pro PreviewGoogle · mean@8 · defaultStrict
|
Clean score95.5% | Cost$18.87 | Clean change+2.9 ppfrom 92.6% | Benchmark sourceStrict |
Evidence detailExpand this record to load its answer-safe evidence detail. | |||||
| 21↑12 |
GPT 6 LunaOpenAI · mean@1 · maxComplete native primary
|
Clean score95.2% | Cost$0.94 | Clean change+4.3 ppfrom 90.9% | Benchmark sourceComplete native primary |
Evidence detailExpand this record to load its answer-safe evidence detail. | |||||
| 21↓1 |
DeepSeek V4.1 FlashDeepSeek · mean@1 · maxComplete native primary
|
Clean score95.2% | Cost$1.36–$2.73 | Clean change+2.8 ppfrom 92.4% | Benchmark sourceComplete native primary |
Evidence detailExpand this record to load its answer-safe evidence detail. | |||||
| 21↓1 |
DeepSeek V4 ProDeepSeek · mean@1 · maxComplete native primary
|
Clean score95.2% | Cost$6.54–$13.08 | Clean change+2.8 ppfrom 92.4% | Benchmark sourceComplete native primary |
Evidence detailExpand this record to load its answer-safe evidence detail. | |||||
| 21↓7 |
Claude Opus 5Anthropic · mean@1 · noneStrict
|
Clean score95.2% | Cost$7.32 | Clean change+2.3 ppfrom 92.9% | Benchmark sourceStrict |
Evidence detailExpand this record to load its answer-safe evidence detail. | |||||
| 21↑4 |
Ember-1Fireworks · mean@1 · maxNative transport composite
|
Clean score95.2% | Cost$10.02 | Clean change+3.3 ppfrom 91.9% | Benchmark sourceNative transport composite |
Evidence detailExpand this record to load its answer-safe evidence detail. | |||||
No configurations match these filters.
All accepted protocols share one standard competition rank ordered by the exact Clean score. Equal exact rational scores share rank, and the next rank skips the tied places. Displayed one-decimal values may coincide; they never create a ranking tie.
The observed rank remains primary. Each expanded record shows its plausible rank interval from one synchronized Record-ID bootstrap shared across configurations.