OpenAIGPT-5.4 Pro
Best configuration ranks #36 of 610 on the BenchLeader Index at 64.2 ±4.0. Last measured 23 Mar 2026. Released 5 Mar 2026.
- Blended price
- $67.50/M
- $30.00 in · $180.00 out
- Output speed
- 1 tok/s
- First answer
- 6.70 s
- Context
- 1.1M
- Overall index64
- Reasoning75
- Knowledge73
- Instruction following67
- Multimodal70
Reasoning-effort configurations
The same model behaves differently depending on how much it is allowed to think. Each row is one setting, scored only on the benchmarks that were run at that setting. “Default” means the publisher did not say which setting was used.
| Effort | Index | Rank | Speed | First answer | Chat reply cost | Categories |
|---|---|---|---|---|---|---|
| no reasoning | – | – | 1 tok/s | 6.70 s | $0.066 | Coding 55 |
| xhigh | 61.1 | #73 | 1 tok/s | 6.70 s | $0.066 | Knowledge 55 · Maths 73 · Reasoning 69 |
| defaultbest | 64.2 | #36 | 1 tok/s | 6.70 s | $0.066 | Instruction following 67 · Knowledge 73 · Multimodal 70 · Reasoning 75 |
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 | no reasoning | xhigh | default | Source |
|---|---|---|---|---|
| GPQA Diamond | – | 94.6%#4 | – | Epoch AI Benchmarking Hub |
| Humanity's Last Exam | – | – | 44.3%#3 | Scale AI / CAIS |
| SimpleBench | – | – | 74.1%#10 | SimpleBench |
| ARC-AGI-1 | – | 94.5%#24 | – | ARC Prize |
| ARC-AGI-2 | – | 83.3%#24 | – | ARC Prize |
Maths
| Benchmark | no reasoning | xhigh | default | Source |
|---|---|---|---|---|
| FrontierMath Tiers 1–3 | – | 82.5%#9 | – | Epoch AI Benchmarking Hub |
| FrontierMath Tier 4 | – | 58.5%#17 | – | Epoch AI Benchmarking Hub |
| MathArena Apex | – | 69.8%#3 | – | MathArena |
Knowledge
| Benchmark | no reasoning | xhigh | default | Source |
|---|---|---|---|---|
| SimpleQA Verified | – | 46.3%#30 | – | Epoch AI Benchmarking Hub |
| MultiNRC | – | – | 62.3%#4 | Scale AI SEAL |
Instruction following
| Benchmark | no reasoning | xhigh | default | Source |
|---|---|---|---|---|
| MultiChallenge | – | – | 69.2%#4 | Scale AI SEAL |
| TutorBench | – | – | 56.6%#2 | Scale AI SEAL |
Multimodal
| Benchmark | no reasoning | xhigh | default | Source |
|---|---|---|---|---|
| VISTA | – | – | 53.9%#2 | Scale AI SEAL |
Composite
| Benchmark | no reasoning | xhigh | default | Source |
|---|---|---|---|---|
| Epoch Capabilities Indexnot in index | – | – | 158.9#9 | Epoch AI Benchmarking Hub |
Where it wins
Benchmarks where this configuration ranks in the top five of every configuration measured.
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 |
|---|---|---|---|---|---|---|
| OpenAI | 1 tok/s | 6.70 s | $30.00 | $180.00 | 1.1M | – |
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 | $0.066 | – | 5.1 min |
| Summarise a 30-page report | 12,000 / 600 | $0.468 | – | 10.1 min |
| Code edit | 6,000 / 1,500 | $0.450 | – | 25.1 min |
| Agentic coding session | 60,000 / 4,000 | $2.52 | – | 66.8 min |
| Structured extraction | 2,000 / 200 | $0.096 | – | 3.4 min |
See also
Data as of 9 Sept 2026. Compare these configurations.