GPT-4
Better value: Gemma 4 E4B scores 43.2 for $0.040/M, 938× cheaper.
197 cheaper ranked models score higher: Claude Fable 5.2 (72.7, $16.00), Gemini 4 Carbon (68.9, $4.00), Claude Opus 5.5 (68.4, $8.00) and 194 more.
As of 15 Nov 2026, GPT-4 by OpenAI scores 42.3 on the BenchLeader Index (608 ranked configurations), with its best results in coding (44) and weakest in reasoning (33). API pricing: $30.00 per 1M input tokens and $60.00 per 1M output tokens ($37.50 blended at 3:1), list price on 15 Nov 2026. It streams about 97 tokens/s (OpenRouter). Released 14 Mar 2023.
Key facts (as of 15 Nov 2026)
- GPT-4 API pricing: $30.00 per 1M input tokens and $60.00 per 1M output tokens ($37.50 blended at 3:1) (list price, 15 Nov 2026).
- GPT-4 streams about 97 tokens/s (OpenRouter, 15 Nov 2026).
- GPT-4 was released on 14 Mar 2023 (OpenAI).
In detail. GPT-4 is an OpenAI proprietary model, released 14 Mar 2023. Its best configuration ranks #496 of 608 on the BenchLeader Index at 42.3 ±6.1, in the lower half. The ± is the point: 220 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 coding (44) and lowest in reasoning (33). At $37.50 per million tokens blended it is among the most expensive ranked models. Output speed of 97 tokens per second puts it faster than most, with a first token in 0.6 s. It has been superseded by gpt-4-0125, which is the version to compare against today. Last measured 11 Nov 2026.
- Blended price
- $37.50/M
- $30.00 in · $60.00 out
- via models.dev
- Output speed
- 97 tok/s
- OpenRouter traffic, 7-day median; not yet measured by Artificial Analysis
- First answer
- 0.59 s
- Context
- 8k
- via models.dev
- Cost per task
- –
- Tokens per task
- –
- Released
- 14 Mar 2023
- Overall index42
- Reasoning33
- Coding44
- Maths38
- Instruction following37
- Human preference44
Versions
OpenAI has shipped 13 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-5.6 | 9 Jul 2026 | – | – |
| GPT-5.5xhigh | 23 Apr 2026 | 64.1 | #31 |
| GPT-5.4xhigh | 5 Mar 2026 | 63.1 | #42 |
| GPT-5.2xhigh | 11 Dec 2025 | 59.9 | #95 |
| GPT-5.1high | 12 Nov 2025 | 57.5 | #151 |
| GPT-5medium | 7 Aug 2025 | 59.0 | #113 |
| GPT-4.1 | 14 Apr 2025 | 48.7 | #377 |
| GPT-4.5 | 27 Feb 2025 | 50.5 | #335 |
| GPT 4 0125 | 25 Jan 2024 | – | – |
| GPT-4 1106 | 6 Nov 2023 | – | – |
| GPT 4 0613 | 13 Jun 2023 | 39.5 | #541 |
| GPT-4this page | 14 Mar 2023 | 42.3 | #496 |
| GPT-3.5 | – | – | – |
How its index has moved
8 Sep 2026 to 15 Nov 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 | 35.7%#271 | Epoch AI Benchmarking Hub | |
| Arena Hard Prompts | 1306#266 | Arena (LMArena) | |
| GPQA Diamond (AA)not in index | 35.0%#502 | Artificial Analysis | |
| Arena Expertnot in index | 1265#275 | Arena (LMArena) | |
| HellaSwagnot in index | 95.3%#1 | Epoch AI Benchmarking Hub | |
| WinoGrandenot in index | 87.5%#3 | Epoch AI Benchmarking Hub |
Coding
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| Arena Coding | 1330#271 | Arena (LMArena) |
Agents & tools
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| GDPval-AA v2.1not in index | -13#288 | Artificial Analysis | |
| METR Time Horizonsnot in index | 36.1%#33 | METR |
Maths
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| OTIS Mock AIME | 0.6%#284 | Epoch AI Benchmarking Hub | |
| Arena Maths | 1285#249 | Arena (LMArena) | |
| GSM8Knot in index | 92.0%#4 | Epoch AI Benchmarking Hub |
Knowledge
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| MMLUnot in index | 86.4%#5 | Epoch AI Benchmarking Hub |
Instruction following
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| IFBench | 33.2%#336 | Artificial Analysis | |
| Arena Instruction Followingnot in index | 1285#262 | Arena (LMArena) |
Human preference
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| Arena Text | 1288#276 | Arena (LMArena) | |
| EQ-Bench Creative Writing v3not in index | 750#117 | EQ-Bench | |
| Arena Creative Writingnot in index | 1255#272 | Arena (LMArena) | |
| Arena Multi-turnnot in index | 1274#275 | Arena (LMArena) | |
| Arena Longer Queriesnot in index | 1280#284 | Arena (LMArena) |
Composite
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| Epoch Capabilities Indexnot in index | 121.8#163 | Epoch AI Benchmarking Hub | |
| AA Intelligence Index v4.3.2not in index | 6.7#520 | Artificial Analysis | |
| AA Progress Index v1not in index | 140.1#573 | Artificial Analysis |
API pricing: 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 | 99 tok/s | 0.81 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 | – | 3.7 s |
| Summarize a 30-page report | 12,000 / 600 | $0.396 | – | 6.8 s |
| Code edit | 6,000 / 1,500 | $0.270 | – | 16.1 s |
| Agentic coding session | 60,000 / 4,000 | $2.04 | – | 41.8 s |
| Structured extraction | 2,000 / 200 | $0.072 | – | 2.7 s |
More from OpenAI
See also
Embed or cite
Cite this
The ranking changes daily, so a reference carries the data date. APA style:
BenchLeader. (2026, November 15). GPT-4 on the BenchLeader Index (method v1.4) [Data set]. Retrieved November 15, 2026, from https://www.benchleader.com/models/gpt-4
BibTeX:
@misc{benchleader_gpt_4_20261115,
author = {{BenchLeader}},
title = {{GPT-4 on the BenchLeader Index (method v1.4)}},
year = {2026},
month = nov,
howpublished = {\url{https://www.benchleader.com/models/gpt-4}},
note = {Data as of 2026-11-15}
}Every badge style, the embeddable top-10 widget and how to cite the Index
Questions people ask
- What are GPT-4's benchmark scores?
- As of 15 Nov 2026, GPT-4 scores 42.3 ±6.1 on the BenchLeader Index, strongest in coding (44) and weakest in reasoning (33). That rests on independent benchmark results only, and every score on this page links to its publisher.
- How does GPT-4 rank?
- GPT-4 is #496 of 608 ranked configurations on 15 Nov 2026. 220 other configurations score within its margin, so the exact position matters less than the band. The leader is Claude Fable 5.2 (max effort) at 72.7.
- How much does GPT-4 cost per 1M tokens?
- As of 15 Nov 2026, GPT-4 costs $30.00 per 1M input tokens and $60.00 per 1M output tokens, $37.50 blended (three input tokens per output token).
- When was GPT-4 released?
- OpenAI released GPT-4 on 14 Mar 2023.
Data as of 15 Nov 2026. Compare with another model.
Cite as: BenchLeader, “GPT-4: benchmarks, pricing, speed and rank”, data as of 15 Nov 2026, https://www.benchleader.com/models/gpt-4.