Inference
EN
OpenAI

GPT-5.4 Pro

GPT-5.4 Pro is a reasoning model from OpenAI in the GPT-5.4 family. 12 benchmarks count toward its index, in 4 categories.

  • Closed weights
  • Reasoning
  • Input: text · image · file
  • Released 2026-03
RankedEnough results for an index.
IndexAll categories, out of 100.73.3±8.7Median 51.0
CoverageIndex weight that has results.60%
SpeedOutput tokens per second.21/sMedian 67/s
First tokenSeconds to the first output token.88.0 sMedian 1.3 s
Input / 1MUS dollars per 1M input tokens.$30.00batch $15.00Median $0.60
Output / 1MUS dollars per 1M output tokens.$180.00batch $90.00 Batch price per 1M output tokens.Median $2.55
ContextMaximum tokens in one request.1.05MMedian 262K
EloLMArena rating and rank.—Median 1422

Details

From OpenRouter
API IDThe model ID on OpenRouter.
openai/gpt-5.4-pro
Max outputMaximum output tokens per request.
128K
Knowledge cutoffLast date of training data.
—
ToolsTool calls on OpenRouter.
Yes
JSON outputOutput follows a given JSON schema.
Yes
Cache read / 1MUS dollars per 1M cached tokens.
—
WeightsLink to the published weights.
—
RetiresDate OpenRouter removes the model.
—

Where to run

APIs that serve this model.
3 providers
ProviderTypeMaker, cloud, host or router.InputOutputContext
DigitalOceanCloud$30.00$180.001.05M
Microsoft AzureCloud$30.00$180.001.05M
OpenRouterRouter$30.00$180.001.05M

US dollars per 1M tokens, list price. From models.dev (opens in a new tab) (MIT).

Capabilities

Index per category, out of 100.
Out of 100
AgenticMulti-step tasks with tools.1 counted70.5—
CodingCode writing and repair.None counted——
ReasoningLogic problems and puzzles.4 counted80.48 / 227
MultimodalTasks with images and text.None counted——
KnowledgeFacts and expert knowledge.3 counted67.234 / 268
MultilingualTasks in many languages.None counted——
InstructionTasks with strict prompt rules.None counted——
MathMath problems.4 counted75.313 / 168

Results

12 counted
BenchmarkThe benchmark name.ResultThe score from the source.PlacePlace among results on this benchmark.
GPQA diamond · Epoch AIKnowledge94.6% ±1.68 / 217
FrontierMath-2025-02-28-PrivateMath50.0% ±2.93 / 69
Critical Physics TasksReasoningAbove fitFar above this model's index.30.0%10 / 214
FrontierMath-Tier-4-2025-07-01-PrivateMath37.5% ±7.03 / 55
Humanity's Last ExamKnowledge58.7%8 / 77
Chess PuzzlesReasoning59.0% ±4.910 / 90
ARC-AGI-2 (semi-private)Reasoning83.3%19 / 131
ARC-AGI-1 (semi-private)Reasoning94.5%21 / 126
BrowseCompAgentic89.3%11 / 63
FrontierMath-Tiers-1-3-v2-PrivateMath82.5% ±2.317 / 90
FrontierMath-Tier-4-v2-PrivateMath58.5% ±7.820 / 72
SimpleQA VerifiedKnowledge46.3% ±1.640 / 91
Humanity's Last Exam without toolsKnowledge42.7%25 / 53
ARC-AGI-1 (public eval)Reasoning98.3%8 / 116
ARC-AGI-2 (public eval)Reasoning92.2%12 / 122
International Physics Olympiad 2025 (Theory)Math93.5%—
FrontierScienceKnowledge36.7%—
FrontierScience ResearchKnowledge36.7%—

Sources

BenchLMCC BY-NC 4.0OpenRouterTerms restrict reuseEpoch AICC BYARC Prize FoundationTerms restrict reusemodels.devMIT

History

5 releases
Index74.6 → 73.3
23 Oct 202612 Nov 2026
Output / 1M$180.00 → $180.00
23 Oct 202612 Nov 2026

Questions

How much does GPT-5.4 Pro cost?
$30.00 per 1M input tokens, $180.00 per 1M output tokens.
How good is GPT-5.4 Pro?
Its index is 73.3 out of 100. That is place 15 of 283 ranked models.
What is GPT-5.4 Pro best at?
Reasoning: 80.4 out of 100, place 8 of 227.
How fast is GPT-5.4 Pro?
21 output tokens per second, 88.0 s to the first token.
What is the context window of GPT-5.4 Pro?
1.05M tokens.
Is GPT-5.4 Pro open source?
No. Its weights are not published.
When was GPT-5.4 Pro released?
March 2026.

Badge

GPT-5.4 Pro on Inference360: place 15 of 283, index 73.3
HTML
<a href="https://inference360.ai/models/gpt-5-4-pro"><img src="https://inference360.ai/badges/gpt-5-4-pro.svg" alt="GPT-5.4 Pro on Inference360: place 15 of 283, index 73.3" height="20"></a>
Markdown
[![GPT-5.4 Pro on Inference360: place 15 of 283, index 73.3](https://inference360.ai/badges/gpt-5-4-pro.svg)](https://inference360.ai/models/gpt-5-4-pro)

Compare with

Muse Wonder73.1GPT-5.6 Sol73.0GPT-6 Sol72.8Claude Opus 573.8Claude Fable 574.0

Lower price

Hy472.0 · $1.25/1MGPT-6.1 Terra77.1 · $5.00/1MGemini 4 Carbon82.7 · $10.00/1M

More from OpenAI

All models
GPT-5.464.8GPT-5.3 InstantUnrankedGPT Realtime 1.5UnrankedGPT-5.4 mini54.5GPT-5.4 nano50.9