OpenAIAuto-detectedGPT-4
Best configuration ranks #477 of 610 on the BenchLeader Index at 41.8 ±5.4. Last measured 2 Sept 2026. Released 13 Jun 2023.
- Blended price
- $37.50/M
- $30.00 in · $60.00 out
- Output speed
- 23 tok/s
- First answer
- 0.72 s
- Context
- 8k
- Overall index42
- Reasoning38
- Coding35
- Maths33
- Knowledge48
- Instruction following37
- Human preference48
- Composite37
Benchmark results
One column per reasoning effort. Rank is among every configuration of every model on that benchmark. Hover a score for the run it came from.
Reasoning
| Benchmark | default | Source |
|---|---|---|
| GPQA Diamond | 42.4%#224 | Epoch AI Benchmarking Hub |
| LMArena Hard Prompts | 1312#228 | LMArena |
| GPQA Diamond (AA)not in index | 35.0%#466 | Artificial Analysis |
Coding
| Benchmark | default | Source |
|---|---|---|
| WeirdML | 12.4%#141 | WeirdML |
| LMArena Coding | 1340#232 | LMArena |
| SWE-bench Verified (any scaffold)not in index | 22.4%#60 | SWE-bench |
Maths
| Benchmark | default | Source |
|---|---|---|
| OTIS Mock AIME | 1.1%#246 | Epoch AI Benchmarking Hub |
| MATH Level 5 | 40.0%#60 | Epoch AI Benchmarking Hub |
Knowledge
| Benchmark | default | Source |
|---|---|---|
| AA-Omniscience | -33.8#230 | Artificial Analysis |
Instruction following
| Benchmark | default | Source |
|---|---|---|
| IFBench | 33.2%#316 | Artificial Analysis |
Human preference
| Benchmark | default | Source |
|---|---|---|
| LMArena Text | 1313#223 | LMArena |
Composite
| Benchmark | default | Source |
|---|---|---|
| Epoch Capabilities Indexnot in index | 125.9#149 | Epoch AI Benchmarking Hub |
| AA Intelligence Index | 6.7#428 | Artificial Analysis |
Where to run it
Every provider serving this model through OpenRouter, with throughput and first-token latency measured on live traffic over the last 30 minutes and each provider’s own price. Purple marks the best in each column.
| Provider | Speed | First token | Input $/M | Output $/M | Context | Quantisation |
|---|---|---|---|---|---|---|
| Azure | 35 tok/s | 0.38 s | $30.00 | $60.00 | 8k | – |
| OpenAI | 11 tok/s | 1.05 s | $30.00 | $60.00 | 8k | – |
What a task costs
Estimates from list price, output speed and time to first answer for the best configuration. “With caching” assumes three-quarters of the input is served from the prompt cache. Reasoning tokens are not modelled.
| Workload | Tokens in / out | Cost | With caching | Time |
|---|---|---|---|---|
| Chat reply | 400 / 300 | $0.030 | – | 13.9 s |
| Summarise a 30-page report | 12,000 / 600 | $0.396 | – | 27.0 s |
| Code edit | 6,000 / 1,500 | $0.270 | – | 1.1 min |
| Agentic coding session | 60,000 / 4,000 | $2.04 | – | 2.9 min |
| Structured extraction | 2,000 / 200 | $0.072 | – | 9.5 s |
See also
Data as of 9 Sept 2026. Compare with another model.