GPT-5 Mini
Aug 2025
GPT-5 MiniCompare inon
Benchmarks of GPT-5 Mini62
BenchmarkByScoreRank
Overall1
Knowledge4
Reasoning4
Math14
Science3
Coding3
Agents4
Multimodal2
Writing17
Health & law4
Documents2
Legacy · saturated or no longer reported4
No benchmark matches
Rankingat high effort?
Alternatives?
ModelScoreCostPer taskWhat changes
CheaperMiMo v2.6 ProunstatedOpenwins 67%-21%$0.43 / $0.87Pass back reasoning_content+3BetterGPT-6 Lunamaxwins 52%+27%$0.068Function calling only with none · Chat Completions+3CheaperMiMo v2.5 ProunstatedOpenwins 64%-21%$0.43 / $0.87No image inputCheaperMiMo v2.6 FlashunstatedOpenwins 63%-75%$0.14 / $0.28Pass back reasoning_content+3
CheaperGLM 5.3 FlashunstatedOpenwins 63%-65%$0.15 / $0.50Defaults to max · Self-hosted+4ModelScoreCostPer taskWhat changes
MiMo v2.6 ProunstatedOpenwins 67%-21%$0.43 / $0.87Pass back reasoning_content+3MiMo v2.5 ProunstatedOpenwins 64%-21%$0.43 / $0.87No image inputMiMo v2.6 FlashunstatedOpenwins 63%-75%$0.14 / $0.28Pass back reasoning_content+3GLM 5.3 FlashunstatedOpenwins 63%-65%$0.15 / $0.50Defaults to max · Self-hosted+4
Qwen3.7 Plusunstatedwins 61%-19%$0.32 / $1.28ModelScoreCostPer taskWhat changes
GPT-6 Lunamaxwins 52%+27%$0.068Function calling only with none · Chat Completions+3
Claude Haiku 5.5highwins 54%+48%$0.079Remove sampling parameters+5GPT-5.6 Lunaxhighwins 59%+60%$0.085GPT-5.6 Terraxhighwins 60%+1,081%$0.63Prompts with >272K input tokensGPT-6.1 Solmaxwins 60%+1,254%$0.72Remove unsupported parameters+5ModelScoreCostPer taskWhat changes
MiMo v2.6 ProunstatedOpenwins 67%-21%$0.43 / $0.87Pass back reasoning_content+3GLM 5.3maxOpenwins 64%+3,649%$2.01No image input+4
Kimi K3maxOpenwins 64%+3,639%$2.00Sampling parameters are fixed+3MiMo v2.5 ProunstatedOpenwins 64%-21%$0.43 / $0.87No image inputGLM 5.1unstatedOpenwins 64%+213%$1.40 / $4.40Context: 400K → 203K+1ModelScoreCostPer taskWhat changes
GPT-5.5highwins 63%+2,516%$1.40Prompts with >272K input tokensGPT-5.4highwins 62%—$2.50 / $15GPT-5.6 Solxhighwins 61%+2,114%$1.18GPT-5.1highwins 60%—$1.25 / $10GPT-6.1 Solmaxwins 60%+1,254%$0.72Remove unsupported parameters+5GPT-5.6 Terraxhighwins 60%+1,081%$0.63Prompts with >272K input tokensGPT-6 Astramaxwins 60%+5,989%$3.26Monitoring can return HTTP 403+2GPT-5.6 Lunaxhighwins 59%+60%$0.085All OpenAI models →History
7 versionsBenchmarks
all categoriesChat & Knowledge
Humanity's Last Exam21.5#120
Arena Text · overall1,390#106
Arena Text · Chinese1,385.4#121
Arena Text · English1,378.9#120
Arena Text · French1,386.3#105
Arena Text · German1,365.8#96
Arena Text · instruction following1,356.7#112
Arena Text · Japanese1,341.2#82
Arena Text · Korean1,306.5#101
Arena Text · non-English1,362.5#111
Arena Text · Polish1,378.4#88
Arena Text · Russian1,361.9#113
Arena Text · Spanish1,354.6#118
AA Omniscience Index-17.3#116
CritPt (physics)0#181
AA Intelligence Index 4.3.216.8#137
AA-LCR (long context) 1.172.3#108
GDP.pdf · AA all-pass8.4#68
MathArena · CMIMC 202584.4#10
MLCR-AA8.9#61
Arena Text · overall · October 22, 20261,373.2#108
Arena Text · overall · October 15, 20261,373.3#104
GPQA Diamond82.8#99
AIME 2025 · AA archived90.7#14
MathArena · AIME 202587.5#16
Coding
SciCode39#101
Terminal-Bench 2.13.7#140
LiveCodeBench · AA archived83.8#10
Agents & Tool Use
AutomationBench-AA6.5#81
GDPval-AA (work tasks) 2.1765.6#90
τ³-Banking · AA15.5#67
AA-Briefcase 1.1 · Elo428.3#102
τ²-bench · AA68.4#95
Writing & Language
Arena Text · creative writing1,323#116
IFBench (instructions)75.4#23
Vision & Documents
MMMU-Pro (vision)70.1#70
Arena Vision · overall1,202.2#65
FAQ
How much does GPT-5 Mini cost?›
$0.25 per 1M input tokens and $2 per 1M output tokens; cached input $0.025 per 1M. A typical benchmark task costs about $0.053 at high effort.
Is GPT-5 Mini free to use?›
There is no free API; the cheapest paid price is $0.25 per 1M input tokens.
How good is GPT-5 Mini, and what is it best at?›
#120 of 336 in Chat & Knowledge, #77 of 134 in Coding, #81 of 135 in Agents & Tool Use, #114 of 197 in Writing & Language, #69 of 125 in Vision & Documents. Strongest at: Instructions.
Is GPT-5 Mini better than GPT-4.1 Mini?›
On Arena Text · English, GPT-5 Mini scores 1,378.9 vs 1,355.8 — higher.
How can I use GPT-5 Mini?›
Through the API of Azure, OpenAI.
What is the context window of GPT-5 Mini?›
400K tokens, with up to 128K output tokens.
When was GPT-5 Mini released? Is there a newer version?›
Released 2025-08-07. It has been superseded by GPT-5.6 Terra.
Is GPT-5 Mini open source?›
No, GPT-5 Mini is a closed model available through an API. The closest open-weights option here is MiMo v2.6 Pro.
Can GPT-5 Mini read images?›
It accepts text and images, PDF files as input.
What is GPT-5 Mini?›
GPT-5 Mini is a closed language model from OpenAI, released Aug 2025, the successor to GPT-4.1 Mini.
What are the best alternatives to GPT-5 Mini?›
MiMo v2.6 Pro (cheaper), GPT-6 Luna (better).