OpenAIGPT-4 Turbo
GPT-4 Turbo is an OpenAI proprietary model, released 9 Apr 2024. Its best configuration ranks #646 of 785 on the BenchLeader Index at 40.9 ±4.6, in the lower half. The ± is the point: 282 other configurations score within that range, so they and this one cannot be told apart on quality alone — price and speed are what separate them. It scores highest in human preference (48) and lowest in reasoning (36). At $15.00 per million tokens blended it is among the most expensive ranked models. Output speed of 33 tokens per second puts it in the slowest quarter, with a first answer in 3.0 s. It has been measured at 2 reasoning-effort settings; this summary describes the best-scoring one, and the tabs above switch between them. Last measured 22 Oct 2026.
- Blended price
- $15.00/M
- $10.00 in · $30.00 out
- Output speed
- 33 tok/s
- measured by Artificial Analysis
- First answer
- 3.00 s
- first token 3.00 s
- Context
- 128k
- Full answer
- 18 s
- median, reasoning included
- Cost per run
- –
- Released
- 9 Apr 2024
- Overall index41
- Reasoning36
- Coding37
- Maths36
- Human preference48
- Multimodal37
- Composite37
Versions
OpenAI has shipped 4 models under this name. Each is ranked on its own results; a newer version often has fewer results so far, which holds its index nearer the average until more arrive.
| Model | Released | Index | Rank |
|---|---|---|---|
| GPT-4 Turbothis page | 9 Apr 2024 | 40.9 | #646 |
| GPT 3.5 Turbo 0125 | 25 Jan 2024 | 35.1 | #765 |
| GPT 3.5 Turbo 1106 | 6 Nov 2023 | 38.7 | #706 |
| GPT-3.5 Turbohigh | 18 Sept 2023 | – | – |
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. “Not stated” collects results from publishers that did not say which setting they used; for a reasoning model that is usually its thinking mode, but we do not assume it. Pick a setting here or at the top of the page to see its price, speed and category scores.
| Effort | Index | Rank | Speed | First answer | Chat reply cost | Categories |
|---|---|---|---|---|---|---|
| high | – | – | 33 tok/s | 3.00 s | $0.013 | Knowledge 51 |
| not statedbest | 40.9 | #646 | 33 tok/s | 3.00 s | $0.013 | Coding 37 · Composite 37 · Human preference 48 · Maths 36 · Multimodal 37 · Reasoning 36 |
How its index has moved
8 Sept 2026 to 23 Oct 2026The index is recomputed from scratch every day, so a line moves when a new benchmark result lands, when a publisher revises a score, or when the models it is normalised against change. Early movement usually means the score is still settling.
Benchmark results
See all LLM benchmarks. 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 | not stated | Source | Trend |
|---|---|---|---|
| GPQA Diamond | 46.6%#231 | Epoch AI Benchmarking Hub | |
| SimpleBench | 25.1%#74 | SimpleBench | |
| Arena Hard Prompts | 1316#247 | Arena (LMArena) | |
| Humanity's Last Exam (AA)not in index | 3.1%#603 | Artificial Analysis | |
| LMCAnot in index | 9.8%#179 | Epoch AI Benchmarking Hub | |
| DTBenchnot in index | 61.6%#178 | Epoch AI Benchmarking Hub | |
| ForecastBenchnot in index | 63.2%#46 | ForecastBench |
Coding
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| WeirdML | 18.0%#144 | WeirdML | |
| Arena Coding | 1347#246 | Arena (LMArena) |
Agents & tools
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| METR Time Horizonsnot in index | 36.7%#31 | METR |
Maths
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| FrontierMath Tiers 1–3 | 0.7%#115 | Epoch AI Benchmarking Hub | |
| OTIS Mock AIME | 6.7%#240 | Epoch AI Benchmarking Hub | |
| MATH Level 5 | 46.7%#56 | Epoch AI Benchmarking Hub | |
| Arena Maths | 1303#228 | Arena (LMArena) |
Knowledge
| Benchmark | high | not stated | Source | Trend |
|---|---|---|---|---|
| LegalBench | 80.5%#84 | – | Vals AI | |
| MedQA | 82.0%#65 | – | Vals AI | |
| MMLUnot in indexeffort not stated | – | 81.3%#17 | Epoch AI Benchmarking Hub |
Instruction following
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| Arena Instruction Followingnot in index | 1303#233 | Arena (LMArena) |
Human preference
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| Arena Text | 1325#229 | Arena (LMArena) | |
| Arena Creative Writingnot in index | 1323#193 | Arena (LMArena) | |
| Arena Multi-turnnot in index | 1313#233 | Arena (LMArena) | |
| Arena Longer Queriesnot in index | 1318#242 | Arena (LMArena) |
Multimodal
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| Arena Vision | 1090#128 | Arena (LMArena) |
Composite
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| Epoch Capabilities Indexnot in index | 127.3#156 | Epoch AI Benchmarking Hub | |
| AA Intelligence Index v4.3.2 | 7.0#497 | Artificial Analysis |
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 | $0.013 | – | 12.1 s |
| Summarize a 30-page report | 12,000 / 600 | $0.138 | – | 21.2 s |
| Code edit | 6,000 / 1,500 | $0.105 | – | 48.5 s |
| Agentic coding session | 60,000 / 4,000 | $0.720 | – | 2.1 min |
| Structured extraction | 2,000 / 200 | $0.026 | – | 9.1 s |
See also
Questions people ask
- What are GPT-4 Turbo's benchmark scores?
- As of 24 Oct 2026, GPT-4 Turbo scores 40.9 ±4.6 on the BenchLeader Index, strongest in human preference (48) and weakest in reasoning (36). That rests on 12 independent benchmark results, and every score on this page links to its publisher.
- How does GPT-4 Turbo rank?
- GPT-4 Turbo is #646 of 785 ranked configurations on 24 Oct 2026. 282 other configurations score within its margin, so the exact position matters less than the band. The leader is Claude Opus 5.5 (max effort) at 72.6.
- How much does GPT-4 Turbo cost per 1M tokens?
- As of 24 Oct 2026, GPT-4 Turbo costs $10.00 per 1M input tokens and $30.00 per 1M output tokens, $15.00 blended (three input tokens per output token).
- When was GPT-4 Turbo released?
- OpenAI released GPT-4 Turbo on 9 Apr 2024. BenchLeader has results for it at 2 reasoning-effort settings.
Data as of 24 Oct 2026. Compare these configurations.
Cite as: BenchLeader, “GPT-4 Turbo: benchmarks, pricing, speed and rank”, data as of 24 Oct 2026, https://www.benchleader.com/models/gpt-4-turbo.