GPT-5 Mini is a compact variant of GPT-5, designed for lighter-weight reasoning tasks with reduced latency and cost. As the successor to o4-mini, it retains GPT-5's instruction-following and safety capabilities while being optimized for everyday workloads that don't require the full model's depth. It offers a practical balance of reasoning ability and efficiency for applications where responsiveness and cost management are key considerations.
OpenAI FreeOpenAI GoOpenAI PlusOpenAI ProAPI|VisionReasoningWeb SearchFile|Proprietary Model
Knowledge Cutoff
2024-05-31
Input → Output Format
Context Memory
400KIN128KOUT
AI Performance Evaluation
Arena Overall Score
1390
±5As of 2026-05-01
Overall Rank
No.121
27,053 Votes
Arena by Ability
Hard Prompts
1403±6No.128
Expert Knowledge
1403±18No.119
Instruction Following
1374±8No.128
Conversation Memory
1372±9No.139
Creative
1325±10No.154
Coding
1429±9No.132
Math
1406±15No.99
Arena by Occupation
Creative Writing
1353±8No.143
Social Sciences
1396±10No.134
Media
1349±9No.128
Business
1380±9No.131
Healthcare
1401±16No.129
Legal
1394±15No.126
Software
1415±7No.134
Mathematics
1409±17No.107
Source:Arena Intelligence
Overall
AA Intelligence Index
21%↓19%
LiveBench
61%↑0%
ForecastBench
60%↑1%
Reasoning & Math
AA Math Index
47%↓28%
GPQA Diamond
69%↓13%
HLE
5.0%↓13%
MMLU-Pro
78%↓4%
AIME 2025
47%↓28%
LB Reasoning
59%↓10%
LB Math
74%↑0%
LB Data
50%↓4%
Coding
AA Coding Index
22%↓15%
LiveCodeBench
55%↓11%
LB Coding
76%↑3%
LB Agentic
35%↓10%
TAU2
32%↓48%
TerminalBench
14%↓20%
SciCode
37%↓5%
Language & Instructions
IFBench
46%↓17%
AA-LCR
36%↓26%
Hallucination (HHEM)
13%↑3%
Factual (HHEM)
87%↓3%
LB Language
69%↓3%
LB IF
64%↑13%
Output Speed
Standard Mode
83tok/s↑6
First Output 0.83s
Reasoning Mode
88tok/s↑1
First Output 48.03s
