Skip to main content
leaderboard
.cn
Toggle Sidebar
Search models or organizations…
⌘K
Explore
Frontier Models
Historical Trends
Local Models
Compare Models
Ranking Studio
Library
Models
Pricing
Organizations
Benchmarks
Data sources
Settings
Toggle Sidebar
leaderboard
.cn
Loading page content
Feedback
Loading page content
Back to models
GPT-5.4
Closed
Share
Add to comparison
Tool use · Reasoning
Overview
Pricing
Evaluations
Evaluation overview
Overall rank
#28
86.4 points · 201 models
Research rank
#19
93.6 points · 195 models
Data coverage
94 items
14 sources · updated through 2026-09-02
Domain performance
Coding
#39 / 181
79.4
Research
#19 / 195
93.6
Agents
#26 / 147
84.4
Vision
#17 / 81
83.8
Long context
#26 / 95
82.6
Evaluation highlights
All evaluations
Arena Agent Tool Hallucination
Agents · Arena.ai Agent · High mode
1.33%
1
/ 52
AIME 2026
Reasoning · Kimi K2.6 Technical Report
99.2%
1
/ 29
Terminal Bench 2.0 (Terminus-2)
Coding · DeepSeek-V4-Pro Technical Report
75.1%
1
/ 29
Graphwalks parents 0K-128K (accuracy)
Long context · OpenAI GPT-5.4 mini/nano Release
89.8%
1
/ 18
Arena Vision Rating
Multimodal · Arena.ai Vision · High mode
1,282.68Elo
14
/ 133
CTF Challenge (Internal)
Cybersecurity · OpenAI GPT-5.5 Release
83.7%
5
/ 5
Basic information
Context (input / output)
1.05M / 128K
Release date
2026-03
Input modalities
Text · Image
Output modalities
Text
Reliable knowledge cutoff
2025-08
Availability
OpenAI
Full pricing
Input
$2.5 /1M
Output
$15 /1M
Related links
Official site
Version relationships
GPT-5.4 family
GPT-5.4
Current
2026-03 · Current model
GPT-5.4 Pro
2026-03 · Same series
GPT-5.4 mini
2026-03 · Same series
GPT-5.4 nano
2026-03 · Same series
Provider variants
High / Codex Harness · Search
Runtime modes
High