← All models
OpenAI · proprietary
GPT-5.4
GPT-5.4 by OpenAI appears in 2 sources with Reasoning at 57.57. Best read for Coding, High intelligence, LLM.
CreatorOpenAI
Release date2026-03-05
Knowledge cutoffNot published
Context1M tokens
Input price$2.5/M tokens
Output price$15/M tokens
Modalitytext + vision
CountryUS
Metrics
All source-backed metrics
LLM Stats Rank6ranking · source_rank
Reasoning57.57reasoning · index_reasoning
Math46.72math · index_math
Coding42.99coding · index_code
Research31.52research · index_search
Writing35.51writing · index_communication
Vision37.07multimodal · index_vision
Tool calling35.02tool_calling · index_tool_calling
Long context27.2long_context · index_long_context
Finance24.29domain · index_finance
Healthcare43.13domain · index_healthcare
GPQA92.8 %reasoning · gpqa_score
Code Arena1,726.27 %coding · coding_arena_score
Humanity Last Exam39.8 %reasoning · hle_score
ARC-AGI v273.3 %reasoning · arc_agi_v2_score
MMMU-Pro81.2 %multimodal · mmmu_pro_score
BrowseComp82.7 %research · browsecomp_score
Toolathlon54.6 %tool_calling · toolathlon_score
MCP Atlas67.2 %tool_calling · mcp_atlas_score
FrontierMath47.6 %math · frontiermath_score
SWE-bench Pro57.7 %coding · swe_bench_pro_score
Context1,000,000 tokenscontext · context
Speed202.98 c/sperformance · throughput
Latency948.55 msperformance · latency
Input price2.5 $/Mpricing · input_price
Output price15 $/Mpricing · output_price
Arena Rating1,454.13metric
Arena Rank18metric
Vote Count20,253metric
Evidence
Citations and source overlap
FAQ
How should I read this profile?
Treat this as a source-backed model dossier, not an EvalKit-run verification. The public values are replicated from linked sources and kept source-scoped.
Is GPT-5.4 verified by EvalKit?
No. EvalKit currently shows 0 verified rows until real run evidence exists.
Why can metrics disagree?
Different sources test different tasks, dates, prompts, and aggregation methods. EvalKit keeps those differences visible instead of merging them into a fake universal score.