OpenAI, proprietary

# GPT-5.4

> GPT-5.4 by OpenAI, released March 2026. Ranked #16 of 354 with a Noometry Index of 59.4. API: $2.50 in / $15 out per M tokens. 1.05M context. Scores, sources and comparisons.
- Canonical page: https://noometry.com/models/gpt-5-4
- Last updated: 2026-10-10
- Title: GPT-5.4 Benchmarks, Price & Rank (October 2026) | Noometry

GPT-5.4 by OpenAI ranks 16th of 354 ranked models on the Noometry Index as of October 2026, with a score of 59.4. Its strongest category is long context, where it ranks 8th. API pricing starts at $2.50 per million input tokens and $15 per million output tokens, with a 1.05M-token context window.

Last verified October 10, 2026

## Specifications

- **Noometry rank:** #16 of 354
- **Index score:** 59.4
- **Evidence:** Confirmed 68 results
- **Provider:** [OpenAI](https://noometry.com/providers/openai)
- **Released:** March 5, 2026
- **Weights:** Proprietary
- **Reasoning:** Yes
- **Context window:** 1.05M
- **Max output:** 128K
- **Input price:** $2.50 / M
- **Output price:** $15 / M
- **Blended price:** $5.63 / M
- **Output speed:** 12 tokens/s [Kagi](https://help.kagi.com/kagi/ai/llm-benchmark.html)
- **Value:** #183 of 219
- **Knowledge cutoff:** August 2025
- **Input:** text, image, pdf

## Category scores

Each category score combines every public result we have in that category.

GPT-5.4 category scores

1.  Coding 52.6
2.  Agentic & Tool Use 46.5
3.  Reasoning 61.8
4.  Math 73.5
5.  Knowledge 65.3
6.  Multimodal 43.7
7.  Multilingual 56.2
8.  Instruction Following 77.1
9.  Long Context 50.3
10.  Writing & Preference 71.9
11.  304050607080

GPT-5.4 category ranks
| Category | Score | Rank | Results |
| --- | --- | --- | --- |
| [Coding](https://noometry.com/best/coding) | 52.6 | #33 | 8 |
| [Agentic & Tool Use](https://noometry.com/best/agentic) | 46.5 | #13 | 6 |
| [Reasoning](https://noometry.com/best/reasoning) | 61.8 | #19 | 13 |
| [Math](https://noometry.com/best/math) | 73.5 | #19 | 6 |
| [Knowledge](https://noometry.com/best/knowledge) | 65.3 | #14 | 5 |
| [Multimodal](https://noometry.com/best/multimodal) | 43.7 | #20 | 3 |
| [Multilingual](https://noometry.com/best/multilingual) | 56.2 | #23 | 1 |
| [Instruction Following](https://noometry.com/best/instruction-following) | 77.1 | #27 | 1 |
| [Long Context](https://noometry.com/best/long-context) | 50.3 | #8 | 3 |
| [Writing & Preference](https://noometry.com/best/writing) | 71.9 | #17 | 5 |

## Strengths and weaknesses

Categories where GPT-5.4 places highest and lowest among the models ranked in each, with its score against that category's median.

### Strongest categories

GPT-5.4: strongest categories
| Category | Score | vs median | Rank |
| --- | --- | --- | --- |
| [Long Context](https://noometry.com/best/long-context) | 50.3 | +9.4 | #8 of 296, top 3% |
| [Knowledge](https://noometry.com/best/knowledge) | 65.3 | +28.0 | #14 of 314, top 5% |
| [Reasoning](https://noometry.com/best/reasoning) | 61.8 | +38.2 | #19 of 350, top 6% |

### Weakest categories

GPT-5.4: weakest categories
| Category | Score | vs median | Rank |
| --- | --- | --- | --- |
| [Multimodal](https://noometry.com/best/multimodal) | 43.7 | +5.1 | #20 of 128, top 16% |
| [Coding](https://noometry.com/best/coding) | 52.6 | +13.9 | #33 of 340, top 10% |
| [Instruction Following](https://noometry.com/best/instruction-following) | 77.1 | +5.9 | #27 of 305, top 9% |

## Closest competitors

The models ranked just above and below GPT-5.4. When scores are this close, price and speed are often the better way to choose.

Models ranked closest to GPT-5.4
| Model | Rank | Score | Blended $/M | Speed |  |
| --- | --- | --- | --- | --- | --- |
| [GPT-6 Sol](https://noometry.com/models/gpt-6-sol) | #12 | 61.8 | $4 | — | [Compare](https://noometry.com/compare/gpt-5-4-vs-gpt-6-sol) |
| [Claude Opus 4.8](https://noometry.com/models/claude-opus-4-8) | #13 | 60.7 | $10 | 34 | [Compare](https://noometry.com/compare/claude-opus-4-8-vs-gpt-5-4) |
| [Gemini 3.7 Flash](https://noometry.com/models/gemini-3-7-flash) | #14 | 59.8 | $1.50 | — | [Compare](https://noometry.com/compare/gemini-3-7-flash-vs-gpt-5-4) |
| [Kimi K3](https://noometry.com/models/kimi-k3) | #15 | 59.5 | $6 | — | [Compare](https://noometry.com/compare/gpt-5-4-vs-kimi-k3) |
| [GPT-5.6 Terra](https://noometry.com/models/gpt-5-6-terra) | #17 | 59.2 | $4.50 | 11 | [Compare](https://noometry.com/compare/gpt-5-4-vs-gpt-5-6-terra) |
| [GPT-5.4 Pro](https://noometry.com/models/gpt-5-4-pro) | #18 | 58.9 | $67.50 | — | [Compare](https://noometry.com/compare/gpt-5-4-vs-gpt-5-4-pro) |
| [Claude Opus 4.7](https://noometry.com/models/claude-opus-4-7) | #19 | 58.3 | $10 | 33 | [Compare](https://noometry.com/compare/claude-opus-4-7-vs-gpt-5-4) |
| [Claude Opus 4.6](https://noometry.com/models/claude-opus-4-6) | #20 | 58.2 | $10 | 19 | [Compare](https://noometry.com/compare/claude-opus-4-6-vs-gpt-5-4) |

Sponsored placements are available on pages like this one. [Advertise on Noometry](https://noometry.com/advertise)

## Benchmark results

Every published result we track, with its source. Bold rows are the ones used for ranking; where several exist we prefer independent runs over self-reported numbers.

### Coding

GPT-5.4 Coding benchmark results
| Benchmark | Score | Position | Setting | Source | Date |
| --- | --- | --- | --- | --- | --- |
| [SWE-bench Verified](https://noometry.com/benchmarks/swe-bench-verified) | 76.9% | #8 of 32, top 25% | high | [Epoch AI](https://epoch.ai/benchmarks) | 2026-03-06 |
| [DeepSWE](https://noometry.com/benchmarks/deepswe) | 51.8% | #23 of 29, top 80% | xhigh | [Epoch AI](https://epoch.ai/benchmarks) |  |
| [LMArena WebDev](https://noometry.com/benchmarks/arena-webdev) | 1465 | #54 of 113, top 48% |  | [LMArena](https://lmarena.ai/leaderboard/webdev) | 2026-10-08 |
| [LMArena WebDev](https://noometry.com/benchmarks/arena-webdev) | 1400 |  |  | [LMArena](https://lmarena.ai/leaderboard/webdev) | 2026-10-08 |
| [LMArena WebDev](https://noometry.com/benchmarks/arena-webdev) | 1444 |  |  | [LMArena](https://lmarena.ai/leaderboard/webdev) | 2026-10-08 |
| [SciCode](https://noometry.com/benchmarks/scicode) | 56.6% | #18 of 121, top 15% | xhigh | [Epoch AI](https://epoch.ai/benchmarks) |  |
| [GSO](https://noometry.com/benchmarks/gso-bench) | 25.5% |  | high | [Epoch AI](https://epoch.ai/benchmarks) |  |
| [GSO](https://noometry.com/benchmarks/gso-bench) | 31.4% | #10 of 31, top 33% | xhigh | [Epoch AI](https://epoch.ai/benchmarks) |  |
| [WeirdML](https://noometry.com/benchmarks/weirdml) | 57.4% |  | none | [Epoch AI](https://epoch.ai/benchmarks) |  |
| [WeirdML](https://noometry.com/benchmarks/weirdml) | 77.7% | #13 of 119, top 11% | xhigh | [Epoch AI](https://epoch.ai/benchmarks) |  |
| [LMArena Coding](https://noometry.com/benchmarks/arena-coding) | 1497 | #24 of 294, top 9% | high | [LMArena](https://lmarena.ai/leaderboard/text) | 2026-10-08 |
| [MirrorCode](https://noometry.com/benchmarks/mirrorcode) | 15.6% | #7 of 9, top 78% | high | [Epoch AI](https://epoch.ai/benchmarks) | 2026-08-10 |
| [ALE-Bench](https://noometry.com/benchmarks/ale-bench) | 1,607 | #13 of 105, top 13% | high | [Epoch AI](https://epoch.ai/benchmarks) |  |
| [ALE-Bench](https://noometry.com/benchmarks/ale-bench) | 1,521 |  | medium | [Epoch AI](https://epoch.ai/benchmarks) |  |
| [ALE-Bench](https://noometry.com/benchmarks/ale-bench) | 1,086 |  | none | [Epoch AI](https://epoch.ai/benchmarks) |  |
| [AlgoTune](https://noometry.com/benchmarks/algotune) | 1.85 | #3 of 18, top 17% | high | [Epoch AI](https://epoch.ai/benchmarks) |  |

### Agentic & Tool Use

GPT-5.4 Agentic & Tool Use benchmark results
| Benchmark | Score | Position | Setting | Source | Date |
| --- | --- | --- | --- | --- | --- |
| [Terminal-Bench](https://noometry.com/benchmarks/terminal-bench) | 81.8% | #2 of 41, top 5% |  | [Epoch AI](https://epoch.ai/benchmarks) |  |
| [APEX-Agents](https://noometry.com/benchmarks/apex-agents) | 52.4% | #23 of 49, top 47% |  | [Epoch AI](https://epoch.ai/benchmarks) |  |
| [τ²-bench Banking](https://noometry.com/benchmarks/tau2-banking) | 39.4% | #10 of 26, top 39% | xhigh | [τ²-bench](https://taubench.com/) | 2026-03-25 |
| [DeepResearch Bench](https://noometry.com/benchmarks/deepresearch-bench) | 35.1% | #24 of 24, top 100% | low | [Epoch AI](https://epoch.ai/benchmarks) |  |
| [PostTrainBench](https://noometry.com/benchmarks/posttrainbench) | 19% | #11 of 11, top 100% | high | [Epoch AI](https://epoch.ai/benchmarks) |  |
| [GBAEval](https://noometry.com/benchmarks/gbaeval) | 45.1% | #10 of 23, top 44% |  | [Epoch AI](https://epoch.ai/benchmarks) |  |
| [LMArena Search](https://noometry.com/benchmarks/arena-search) | 1197 | #16 of 32, top 50% |  | [LMArena](https://lmarena.ai/leaderboard/search) | 2026-08-24 |
| [METR Time Horizons](https://noometry.com/benchmarks/metr-time-horizons) | 74.3% | #7 of 32, top 22% | xhigh | [Epoch AI](https://epoch.ai/benchmarks) |  |
| [Vending-Bench 2](https://noometry.com/benchmarks/vending-bench-2) | 6,144 | #19 of 60, top 32% |  | [Epoch AI](https://epoch.ai/benchmarks) |  |

### Reasoning

GPT-5.4 Reasoning benchmark results
| Benchmark | Score | Position | Setting | Source | Date |
| --- | --- | --- | --- | --- | --- |
| [ARC-AGI-2](https://noometry.com/benchmarks/arc-agi-2) | 67.5% |  | high | [Epoch AI](https://epoch.ai/benchmarks) |  |
| [ARC-AGI-2](https://noometry.com/benchmarks/arc-agi-2) | 29.2% |  | low | [Epoch AI](https://epoch.ai/benchmarks) |  |
| [ARC-AGI-2](https://noometry.com/benchmarks/arc-agi-2) | 55.4% |  | medium | [Epoch AI](https://epoch.ai/benchmarks) |  |
| [ARC-AGI-2](https://noometry.com/benchmarks/arc-agi-2) | 74% | #18 of 83, top 22% | xhigh | [Epoch AI](https://epoch.ai/benchmarks) |  |
| [Kagi LLM Benchmark](https://noometry.com/benchmarks/kagi-reasoning) | 63.8% | #34 of 99, top 35% |  | [Kagi LLM Benchmark](https://help.kagi.com/kagi/ai/llm-benchmark.html) |  |
| [NYT Connections (extended)](https://noometry.com/benchmarks/nyt-connections) | 91.3% | #15 of 91, top 17% | xhigh reasoning | [Lech Mazur benchmarks](https://github.com/lechmazur/nyt-connections) |  |
| [ARC-AGI-1](https://noometry.com/benchmarks/arc-agi-1) | 92.7% |  | high | [Epoch AI](https://epoch.ai/benchmarks) |  |
| [ARC-AGI-1](https://noometry.com/benchmarks/arc-agi-1) | 68.2% |  | low | [Epoch AI](https://epoch.ai/benchmarks) |  |
| [ARC-AGI-1](https://noometry.com/benchmarks/arc-agi-1) | 86.2% |  | medium | [Epoch AI](https://epoch.ai/benchmarks) |  |
| [ARC-AGI-1](https://noometry.com/benchmarks/arc-agi-1) | 93.7% | #19 of 83, top 23% | xhigh | [Epoch AI](https://epoch.ai/benchmarks) |  |
| [CritPt](https://noometry.com/benchmarks/critpt) | 23.4% | #18 of 134, top 14% | xhigh | [Epoch AI](https://epoch.ai/benchmarks) |  |
| [Chess Puzzles](https://noometry.com/benchmarks/chess-puzzles) | 38% |  | high | [Epoch AI](https://epoch.ai/benchmarks) | 2026-07-15 |
| [Chess Puzzles](https://noometry.com/benchmarks/chess-puzzles) | 20% |  | low | [Epoch AI](https://epoch.ai/benchmarks) | 2026-07-15 |
| [Chess Puzzles](https://noometry.com/benchmarks/chess-puzzles) | 38% |  | medium | [Epoch AI](https://epoch.ai/benchmarks) | 2026-07-15 |
| [Chess Puzzles](https://noometry.com/benchmarks/chess-puzzles) | 5% |  | none | [Epoch AI](https://epoch.ai/benchmarks) | 2026-07-15 |
| [Chess Puzzles](https://noometry.com/benchmarks/chess-puzzles) | 44% | #15 of 129, top 12% | xhigh | [Epoch AI](https://epoch.ai/benchmarks) | 2026-03-11 |
| [EnigmaEval](https://noometry.com/benchmarks/enigmaeval) | 16% | #9 of 38, top 24% | xhigh | [Epoch AI](https://epoch.ai/benchmarks) |  |
| [Thematic Generalization](https://noometry.com/benchmarks/thematic-generalization) | 80% | #2 of 23, top 9% | xhigh reasoning | [Lech Mazur benchmarks](https://github.com/lechmazur/generalization) |  |
| [EBR-Bench](https://noometry.com/benchmarks/ebr-bench) | 25.4% | #12 of 24, top 50% | xhigh | [Epoch AI](https://epoch.ai/benchmarks) | 2026-06-25 |
| [LMArena Hard Prompts](https://noometry.com/benchmarks/arena-hard-prompts) | 1485 | #24 of 297, top 9% | high | [LMArena](https://lmarena.ai/leaderboard/text) | 2026-10-08 |
| [Mystery Game Puzzles](https://noometry.com/benchmarks/mystery-game-puzzles) | 17% |  | low | [Epoch AI](https://epoch.ai/benchmarks) | 2026-08-27 |
| [Mystery Game Puzzles](https://noometry.com/benchmarks/mystery-game-puzzles) | 28% |  | medium | [Epoch AI](https://epoch.ai/benchmarks) | 2026-08-28 |
| [Mystery Game Puzzles](https://noometry.com/benchmarks/mystery-game-puzzles) | 16% |  | none | [Epoch AI](https://epoch.ai/benchmarks) | 2026-08-30 |
| [Mystery Game Puzzles](https://noometry.com/benchmarks/mystery-game-puzzles) | 37% | #16 of 74, top 22% | xhigh | [Epoch AI](https://epoch.ai/benchmarks) | 2026-07-24 |
| [DTBench](https://noometry.com/benchmarks/dtbench) | 94.4% | #22 of 151, top 15% | xhigh | [Epoch AI](https://epoch.ai/benchmarks) |  |
| [LMCA](https://noometry.com/benchmarks/lmca) | 52% | #19 of 125, top 16% | xhigh | [Epoch AI](https://epoch.ai/benchmarks) |  |
| [Epoch Capabilities Index](https://noometry.com/benchmarks/epoch-capabilities-index) | 156.81 | #17 of 213, top 8% |  | [Epoch AI](https://epoch.ai/eci) | 2026-03-05 |
| [ForecastBench](https://noometry.com/benchmarks/forecastbench) | 59.5 | #38 of 72, top 53% |  | [Epoch AI](https://epoch.ai/benchmarks) |  |

### Math

GPT-5.4 Math benchmark results
| Benchmark | Score | Position | Setting | Source | Date |
| --- | --- | --- | --- | --- | --- |
| [FrontierMath (Tiers 1-3)](https://noometry.com/benchmarks/frontiermath) | 78.6% | #17 of 81, top 21% | xhigh | [Epoch AI](https://epoch.ai/benchmarks) | 2026-06-11 |
| [FrontierMath Tier 4](https://noometry.com/benchmarks/frontiermath-tier-4) | 49% | #18 of 63, top 29% | xhigh | [Epoch AI](https://epoch.ai/benchmarks) | 2026-06-11 |
| [MathArena Final-Answer Competitions](https://noometry.com/benchmarks/matharena) | 83.1% | #5 of 29, top 18% | xhigh | [MathArena](https://matharena.ai/) |  |
| [OTIS Mock AIME 2024-2025](https://noometry.com/benchmarks/otis-mock-aime) | 97.8% | #25 of 173, top 15% | high | [Epoch AI](https://epoch.ai/benchmarks) | 2026-07-15 |
| [OTIS Mock AIME 2024-2025](https://noometry.com/benchmarks/otis-mock-aime) | 84.4% |  | low | [Epoch AI](https://epoch.ai/benchmarks) | 2026-07-15 |
| [OTIS Mock AIME 2024-2025](https://noometry.com/benchmarks/otis-mock-aime) | 95.6% |  | medium | [Epoch AI](https://epoch.ai/benchmarks) | 2026-07-15 |
| [OTIS Mock AIME 2024-2025](https://noometry.com/benchmarks/otis-mock-aime) | 57.8% |  | none | [Epoch AI](https://epoch.ai/benchmarks) | 2026-07-15 |
| [OTIS Mock AIME 2024-2025](https://noometry.com/benchmarks/otis-mock-aime) | 95.3% |  | xhigh | [Epoch AI](https://epoch.ai/benchmarks) | 2026-03-06 |
| [ProofBench](https://noometry.com/benchmarks/proofbench) | 56% | #24 of 77, top 32% | xhigh | [Epoch AI](https://epoch.ai/benchmarks) |  |
| [LMArena Math](https://noometry.com/benchmarks/arena-math) | 1488 | #21 of 285, top 8% | high | [LMArena](https://lmarena.ai/leaderboard/text) | 2026-10-08 |
| [FrontierMath (Feb 2025 set)](https://noometry.com/benchmarks/frontiermath-2025-02) | 47.6% | #4 of 68, top 6% | xhigh | [Epoch AI](https://epoch.ai/benchmarks) | 2026-03-06 |
| [FrontierMath Tier 4 (v1)](https://noometry.com/benchmarks/frontiermath-tier-4-v1) | 27.1% | #7 of 55, top 13% | xhigh | [Epoch AI](https://epoch.ai/benchmarks) | 2026-03-06 |

### Knowledge

GPT-5.4 Knowledge benchmark results
| Benchmark | Score | Position | Setting | Source | Date |
| --- | --- | --- | --- | --- | --- |
| [GPQA Diamond](https://noometry.com/benchmarks/gpqa-diamond) | 89.9% |  | high | [Epoch AI](https://epoch.ai/benchmarks) | 2026-07-15 |
| [GPQA Diamond](https://noometry.com/benchmarks/gpqa-diamond) | 84.8% |  | low | [Epoch AI](https://epoch.ai/benchmarks) | 2026-07-15 |
| [GPQA Diamond](https://noometry.com/benchmarks/gpqa-diamond) | 88.9% |  | medium | [Epoch AI](https://epoch.ai/benchmarks) | 2026-07-15 |
| [GPQA Diamond](https://noometry.com/benchmarks/gpqa-diamond) | 74.7% |  | none | [Epoch AI](https://epoch.ai/benchmarks) | 2026-07-15 |
| [GPQA Diamond](https://noometry.com/benchmarks/gpqa-diamond) | 93.3% | #17 of 186, top 10% | xhigh | [Epoch AI](https://epoch.ai/benchmarks) | 2026-03-06 |
| [Humanity's Last Exam](https://noometry.com/benchmarks/hle) | 36.2% | #8 of 41, top 20% | xhigh | [Epoch AI](https://epoch.ai/benchmarks) |  |
| [SimpleQA Verified](https://noometry.com/benchmarks/simpleqa-verified) | 45.1% | #36 of 77, top 47% | xhigh | [Epoch AI](https://epoch.ai/benchmarks) | 2026-08-27 |
| [Vectara Hallucination Rate](https://noometry.com/benchmarks/vectara-hallucination) (lower is better) | 7% | #29 of 96, top 31% |  | [Vectara Hallucination Leaderboard](https://github.com/vectara/hallucination-leaderboard) |  |
| [LMArena Expert](https://noometry.com/benchmarks/arena-expert) | 1507 | #19 of 273, top 7% | high | [LMArena](https://lmarena.ai/leaderboard/text) | 2026-10-08 |

### Multimodal

GPT-5.4 Multimodal benchmark results
| Benchmark | Score | Position | Setting | Source | Date |
| --- | --- | --- | --- | --- | --- |
| [LMArena Vision](https://noometry.com/benchmarks/arena-vision) | 1303 | #15 of 122, top 13% | high | [LMArena](https://lmarena.ai/leaderboard/vision) | 2026-10-09 |
| [Blueprint-Bench 2](https://noometry.com/benchmarks/blueprint-bench-2) | 27.1% | #19 of 31, top 62% |  | [Epoch AI](https://epoch.ai/benchmarks) |  |
| [Furniture Assembly](https://noometry.com/benchmarks/furniture-assembly) | 37.5% | #17 of 31, top 55% | xhigh | [Epoch AI](https://epoch.ai/benchmarks) | 2026-09-10 |
| [LMArena Document](https://noometry.com/benchmarks/arena-document) | 1471 | #11 of 38, top 29% |  | [LMArena](https://lmarena.ai/leaderboard/document) | 2026-09-13 |

### Multilingual

GPT-5.4 Multilingual benchmark results
| Benchmark | Score | Position | Setting | Source | Date |
| --- | --- | --- | --- | --- | --- |
| [LMArena Non-English](https://noometry.com/benchmarks/arena-non-english) | 1465 | #23 of 297, top 8% | high | [LMArena](https://lmarena.ai/leaderboard/text) | 2026-10-08 |
| [LMArena Chinese](https://noometry.com/benchmarks/arena-chinese) | 1519 | #28 of 285, top 10% | high | [LMArena](https://lmarena.ai/leaderboard/text) | 2026-10-08 |
| [LMArena French](https://noometry.com/benchmarks/arena-french) | 1493 | #17 of 223, top 8% | high | [LMArena](https://lmarena.ai/leaderboard/text) | 2026-10-08 |
| [LMArena German](https://noometry.com/benchmarks/arena-german) | 1472 | #23 of 231, top 10% | high | [LMArena](https://lmarena.ai/leaderboard/text) | 2026-10-08 |
| [LMArena Japanese](https://noometry.com/benchmarks/arena-japanese) | 1485 | #13 of 211, top 7% | high | [LMArena](https://lmarena.ai/leaderboard/text) | 2026-10-08 |
| [LMArena Korean](https://noometry.com/benchmarks/arena-korean) | 1448 | #19 of 213, top 9% | high | [LMArena](https://lmarena.ai/leaderboard/text) | 2026-10-08 |
| [LMArena Russian](https://noometry.com/benchmarks/arena-russian) | 1480 | #21 of 283, top 8% | high | [LMArena](https://lmarena.ai/leaderboard/text) | 2026-10-08 |
| [LMArena Spanish](https://noometry.com/benchmarks/arena-spanish) | 1454 | #46 of 226, top 21% | high | [LMArena](https://lmarena.ai/leaderboard/text) | 2026-10-08 |

### Instruction Following

GPT-5.4 Instruction Following benchmark results
| Benchmark | Score | Position | Setting | Source | Date |
| --- | --- | --- | --- | --- | --- |
| [LMArena Instruction Following](https://noometry.com/benchmarks/arena-instruction-following) | 1469 | #24 of 298, top 9% | high | [LMArena](https://lmarena.ai/leaderboard/text) | 2026-10-08 |

### Long Context

GPT-5.4 Long Context benchmark results
| Benchmark | Score | Position | Setting | Source | Date |
| --- | --- | --- | --- | --- | --- |
| [CL-bench](https://noometry.com/benchmarks/cl-bench) | 27.9% | Best of 19 | xhigh | [Epoch AI](https://epoch.ai/benchmarks) |  |
| [CL-bench Life](https://noometry.com/benchmarks/cl-bench-life) | 13.8% |  |  | [Epoch AI](https://epoch.ai/benchmarks) |  |
| [CL-bench Life](https://noometry.com/benchmarks/cl-bench-life) | 19.3% |  | high | [Epoch AI](https://epoch.ai/benchmarks) |  |
| [CL-bench Life](https://noometry.com/benchmarks/cl-bench-life) | 21.7% | #2 of 13, top 16% | xhigh | [Epoch AI](https://epoch.ai/benchmarks) |  |
| [LMArena Longer Query](https://noometry.com/benchmarks/arena-longer-query) | 1473 | #32 of 291, top 11% | high | [LMArena](https://lmarena.ai/leaderboard/text) | 2026-10-08 |

### Writing & Preference

GPT-5.4 Writing & Preference benchmark results
| Benchmark | Score | Position | Setting | Source | Date |
| --- | --- | --- | --- | --- | --- |
| [LMArena Text](https://noometry.com/benchmarks/arena-text) | 1469 | #28 of 297, top 10% | high | [LMArena](https://lmarena.ai/leaderboard/text) | 2026-10-08 |
| [LMArena Creative Writing](https://noometry.com/benchmarks/arena-creative-writing) | 1439 | #40 of 295, top 14% | high | [LMArena](https://lmarena.ai/leaderboard/text) | 2026-10-08 |
| [EQ-Bench Creative Writing](https://noometry.com/benchmarks/eqbench-creative-writing) | 1840 | #20 of 115, top 18% |  | [EQ-Bench](https://eqbench.com/creative_writing.html) |  |
| [EQ-Bench 4](https://noometry.com/benchmarks/eqbench-4) | 1272 | #7 of 28, top 25% |  | [EQ-Bench](https://eqbench.com/) |  |
| [LMArena Multi-Turn](https://noometry.com/benchmarks/arena-multi-turn) | 1482 | #17 of 295, top 6% | high | [LMArena](https://lmarena.ai/leaderboard/text) | 2026-10-08 |

## API pricing by provider

GPT-5.4 API prices
| Route | Input $/M | Output $/M | Cached input $/M | Checked |
| --- | --- | --- | --- | --- |
| [azure](https://learn.microsoft.com/en-us/azure/ai-services/openai/concepts/models) | $2.50 | $15 | $0.25 | 2026-10-10 |
| [bedrock](https://docs.aws.amazon.com/bedrock/latest/userguide/models-supported.html) | $2.75 | $16.50 | $0.28 | 2026-10-10 |
| [openai](https://platform.openai.com/docs/models) | $2.50 | $15 | $0.25 | 2026-10-10 |
| [openrouter](https://openrouter.ai/openai/gpt-5.4) | $2.50 | $15 | $0.25 | 2026-10-10 |

[All OpenAI API prices →](https://noometry.com/llm-pricing/openai) [Estimate your cost →](https://noometry.com/tools/cost-calculator)

## Compare GPT-5.4

-   [GPT-5.4 vs GPT-5.2](https://noometry.com/compare/gpt-5-2-vs-gpt-5-4)
-   [GPT-5.4 vs Kimi K3](https://noometry.com/compare/gpt-5-4-vs-kimi-k3)
-   [GPT-5.4 vs GPT-5.6 Terra](https://noometry.com/compare/gpt-5-4-vs-gpt-5-6-terra)
-   [GPT-5.4 vs Gemini 3.7 Flash](https://noometry.com/compare/gemini-3-7-flash-vs-gpt-5-4)
-   [GPT-5.4 vs GPT-5.4 Pro](https://noometry.com/compare/gpt-5-4-vs-gpt-5-4-pro)
-   [GPT-5.4 vs Claude Opus 4.8](https://noometry.com/compare/claude-opus-4-8-vs-gpt-5-4)
-   [GPT-5.4 vs Claude Opus 4.7](https://noometry.com/compare/claude-opus-4-7-vs-gpt-5-4)
-   [GPT-5.4 vs Claude Fable 5.1](https://noometry.com/compare/claude-fable-5-1-vs-gpt-5-4)
-   [GPT-5.4 vs Gemini 3.8 Flash](https://noometry.com/compare/gemini-3-8-flash-vs-gpt-5-4)
-   [GPT-5.4 vs Grok 4.6](https://noometry.com/compare/gpt-5-4-vs-grok-4-6)
-   [GPT-5.4 vs Qwen3.8 Max](https://noometry.com/compare/gpt-5-4-vs-qwen3-8-max)
-   [GPT-5.4 vs GLM-5.3](https://noometry.com/compare/glm-5-3-vs-gpt-5-4)
-   [GPT-5.4 vs Muse Spark 1.3](https://noometry.com/compare/gpt-5-4-vs-muse-spark-1-3)
-   [GPT-5.4 vs DeepSeek V4 Pro](https://noometry.com/compare/deepseek-v4-pro-vs-gpt-5-4)

## Other OpenAI models

-   [GPT-6 Astra](https://noometry.com/models/gpt-6-astra)70.8
-   [GPT-6.1 Sol](https://noometry.com/models/gpt-6-1-sol)65.6
-   [GPT-5.6 Sol](https://noometry.com/models/gpt-5-6-sol)65.0
-   [GPT-5.5 Pro](https://noometry.com/models/gpt-5-5-pro)64.3
-   [GPT-5.5](https://noometry.com/models/gpt-5-5)63.4
-   [GPT-6 Sol](https://noometry.com/models/gpt-6-sol)61.8
-   [GPT-5.6 Terra](https://noometry.com/models/gpt-5-6-terra)59.2
-   [GPT-5.4 Pro](https://noometry.com/models/gpt-5-4-pro)58.9

## Frequently asked questions

### How good is GPT-5.4?

GPT-5.4 by OpenAI ranks 16th of 354 ranked models on the Noometry Index as of October 2026, with a score of 59.4. Its strongest category is long context, where it ranks 8th. API pricing starts at $2.50 per million input tokens and $15 per million output tokens, with a 1.05M-token context window.

### How much does GPT-5.4 cost?

GPT-5.4 costs $2.50 per million input tokens and $15 per million output tokens on OpenAI's own API, with cached input at $0.25.

### What is GPT-5.4's context window?

GPT-5.4 accepts up to 1.05M tokens of input and can write up to 128K tokens in one response.

### Is GPT-5.4 open source?

No. GPT-5.4 is proprietary and available only through OpenAI's API and partner platforms.

### How fast is GPT-5.4?

GPT-5.4 generated about 12 output tokens per second in the Kagi LLM Benchmark's timed runs. Speed varies by provider, load and reasoning effort.

### What are GPT-5.4's strengths and weaknesses?

Relative to other ranked models, GPT-5.4 places best in long context, knowledge, reasoning and lowest in multimodal, coding, instruction following.

### What is GPT-5.4 best at?

Its best category is long context, where it ranks 8th on Noometry.

### Cite this page

Noometry. (2026). GPT-5.4 benchmarks and pricing. Retrieved October 10, 2026, from https://noometry.com/models/gpt-5-4

Quote Noometry with a link back to this page. It is also available in [Markdown](https://noometry.com/md/models/gpt-5-4.md).