Models / GPT-5 Mini

OpenAI

GPT-5 Mini

25 published results from 3 sources. Each card shows where the number comes from and what it does not measure. Results are never combined into one score.

Provider
OpenAI
Sources
3
Our benchmarks
0
Price
Not yet published

Reported by others

Reported by Arena (formerly LMArena)
1,394
Overall · rank 144 of 177
Unit
Arena rating, higher is better
Range
1,387 to 1,401
Sample
6843 votes
Configuration
GPT-5 Mini (high)
Measured
25 Sep 2026
Not shown
Not an error rate: claims that cannot be checked on the web are skipped, and preference still carries most of the weight.
Reported by Arena (formerly LMArena)
1,381
Business, management and finance · rank 156 of 402
Unit
Arena rating, higher is better
Range
1,372 to 1,390
Sample
4812 votes
Configuration
GPT-5 Mini (high)
Measured
25 Sep 2026
Not shown
Not accuracy or correctness: it ranks which answer voters preferred, with answer length and formatting controlled for.
Reported by Arena (formerly LMArena)
1,324
Creative writing · rank 178 of 407
Unit
Arena rating, higher is better
Range
1,313 to 1,334
Sample
3491 votes
Configuration
GPT-5 Mini (high)
Measured
25 Sep 2026
Not shown
Not accuracy or correctness: it ranks which answer voters preferred, with answer length and formatting controlled for.
Reported by Arena (formerly LMArena)
1,400
Expert prompts · rank 133 of 359
Unit
Arena rating, higher is better
Range
1,383 to 1,418
Sample
1208 votes
Configuration
GPT-5 Mini (high)
Measured
25 Sep 2026
Not shown
Not accuracy or correctness: it ranks which answer voters preferred, with answer length and formatting controlled for.
Reported by Arena (formerly LMArena)
1,373
Instruction following · rank 152 of 409
Unit
Arena rating, higher is better
Range
1,365 to 1,381
Sample
6947 votes
Configuration
GPT-5 Mini (high)
Measured
25 Sep 2026
Not shown
Not accuracy or correctness: it ranks which answer voters preferred, with answer length and formatting controlled for.
Reported by Arena (formerly LMArena)
1,390
Overall · rank 156 of 409
Unit
Arena rating, higher is better
Range
1,385 to 1,394
Sample
26641 votes
Configuration
GPT-5 Mini (high)
Measured
25 Sep 2026
Not shown
Not accuracy or correctness: it ranks which answer voters preferred, with answer length and formatting controlled for.
Reported by Arena (formerly LMArena)
1,352
Writing, literature and language · rank 166 of 408
Unit
Arena rating, higher is better
Range
1,344 to 1,360
Sample
6095 votes
Configuration
GPT-5 Mini (high)
Measured
25 Sep 2026
Not shown
Not accuracy or correctness: it ranks which answer voters preferred, with answer length and formatting controlled for.
Reported by Berkeley Function Calling Leaderboard (BFCL) V4
44.3%
Memory · rank 15 of 109
Unit
% correct, higher is better
Configuration
gpt-5-mini
Measured
16 Dec 2025
Not shown
Not long-term personal memory in a product: sessions are BFCL's scripted ones.
Reported by Berkeley Function Calling Leaderboard (BFCL) V4
29.2%
Memory · rank 22 of 109
Unit
% correct, higher is better
Configuration
gpt-5-mini
Measured
16 Dec 2025
Not shown
Not long-term personal memory in a product: sessions are BFCL's scripted ones.
Reported by Berkeley Function Calling Leaderboard (BFCL) V4
27.5%
Multi-turn tasks · rank 43 of 109
Unit
% correct, higher is better
Configuration
gpt-5-mini
Measured
16 Dec 2025
Not shown
Not open-ended agent work: the tools and tasks are BFCL's simulated APIs.
Reported by Berkeley Function Calling Leaderboard (BFCL) V4
5.5%
Multi-turn tasks · rank 80 of 109
Unit
% correct, higher is better
Configuration
gpt-5-mini
Measured
16 Dec 2025
Not shown
Not open-ended agent work: the tools and tasks are BFCL's simulated APIs.
Reported by Berkeley Function Calling Leaderboard (BFCL) V4
55.5%
Overall accuracy · rank 17 of 109
Unit
% correct, higher is better
Configuration
gpt-5-mini
Measured
16 Dec 2025
Not shown
Not a neutral average: the weighting is BFCL's. Not reliability on your own tools: BFCL's functions and queries are a fixed test set, and a correct call is judged by its form, not by what it achieved.
Reported by Berkeley Function Calling Leaderboard (BFCL) V4
27.8%
Overall accuracy · rank 77 of 109
Unit
% correct, higher is better
Configuration
gpt-5-mini
Measured
16 Dec 2025
Not shown
Not a neutral average: the weighting is BFCL's. Not reliability on your own tools: BFCL's functions and queries are a fixed test set, and a correct call is judged by its form, not by what it achieved.
Reported by Berkeley Function Calling Leaderboard (BFCL) V4
62.5%
Relevance detection · rank 91 of 109
Unit
% correct, higher is better
Configuration
gpt-5-mini
Measured
16 Dec 2025
Not shown
Not whether the call itself was right; only that one was attempted.
Reported by Berkeley Function Calling Leaderboard (BFCL) V4
93.8%
Relevance detection · rank 8 of 109
Unit
% correct, higher is better
Configuration
gpt-5-mini
Measured
16 Dec 2025
Not shown
Not whether the call itself was right; only that one was attempted.
Reported by Berkeley Function Calling Leaderboard (BFCL) V4
69.8%
Single-turn calls (curated) · rank 87 of 109
Unit
% correct, higher is better
Configuration
gpt-5-mini
Measured
16 Dec 2025
Not shown
Not reliability on your own tools: BFCL's functions and queries are a fixed test set, and a correct call is judged by its form, not by what it achieved.
Reported by Berkeley Function Calling Leaderboard (BFCL) V4
68.0%
Single-turn calls (curated) · rank 90 of 109
Unit
% correct, higher is better
Configuration
gpt-5-mini
Measured
16 Dec 2025
Not shown
Not reliability on your own tools: BFCL's functions and queries are a fixed test set, and a correct call is judged by its form, not by what it achieved.
Reported by Berkeley Function Calling Leaderboard (BFCL) V4
58.6%
Single-turn calls (user-contributed) · rank 90 of 109
Unit
% correct, higher is better
Configuration
gpt-5-mini
Measured
16 Dec 2025
Not shown
Not reliability on your own tools: BFCL's functions and queries are a fixed test set, and a correct call is judged by its form, not by what it achieved.
Reported by Berkeley Function Calling Leaderboard (BFCL) V4
62.5%
Single-turn calls (user-contributed) · rank 82 of 109
Unit
% correct, higher is better
Configuration
gpt-5-mini
Measured
16 Dec 2025
Not shown
Not reliability on your own tools: BFCL's functions and queries are a fixed test set, and a correct call is judged by its form, not by what it achieved.
Reported by Berkeley Function Calling Leaderboard (BFCL) V4
82.0%
Web search · rank 4 of 109
Unit
% correct, higher is better
Configuration
gpt-5-mini
Measured
16 Dec 2025
Not shown
Not general research quality: questions have short, checkable answers.
Reported by Berkeley Function Calling Leaderboard (BFCL) V4
8.5%
Web search · rank 58 of 109
Unit
% correct, higher is better
Configuration
gpt-5-mini
Measured
16 Dec 2025
Not shown
Not general research quality: questions have short, checkable answers.
Reported by Vectara Hallucination Leaderboard
99.9%
Answer rate · rank 19 of 108
Unit
% of documents, higher is better
Configuration
GPT-5 Mini
Measured
22 Sep 2026
Not shown
Not a quality score: a low rate usually means content filters were triggered, and hallucination rates are measured on answered documents only.
Reported by Vectara Hallucination Leaderboard
12.9%
Hallucination rate · rank 86 of 108
Unit
% of summaries, lower is better
Configuration
GPT-5 Mini
Measured
22 Sep 2026
Not shown
Not errors in open questions or other tasks: only summarisation, judged by Vectara's own model (HHEM-2.3), not by people, on news-style documents rather than your data.

Compare GPT-5 Mini with