GeekRadar / The comparison desk

GPT-5 Mini vs GPT-5 Nano

Catalog Differences only

Capability & preference

General capability · ECI

A cross-benchmark capability estimate. Index points, not a percentage; intervals show uncertainty.

GPT-5 Mini 145.4^* GPT-5 mini · 95% interval 143.5–147.1

GPT-5 Nano 139.3^* GPT-5 nano · 95% interval 134.3–141.5

Human preference · Arena

Overall text preferences from head-to-head votes. Not factual accuracy or a task-success rate.

GPT-5 Mini 1373.2 GPT-5 mini (High) · 95% interval 1368.6–1377.8 · 26,793 votes · 2026-10-28

GPT-5 Nano 1319.9 GPT-5 nano (High) · 95% interval 1313.0–1326.7 · 8,134 votes · 2026-10-28

Model performance

Radar task score: four tests, equal weight. An experimental summary of these tasks, not a universal intelligence rating.

GPT-5 Mini

Tested configuration Radar task score

Unrated

Older results · no current score

6 benchmark results for this configuration

GPT-5 Nano

Tested configuration Radar task score

Unrated

Older results · no current score

6 benchmark results for this configuration

Scientific reasoning

GPQA Diamond · 25%

Expert-written science questions; strict multiple-choice scoring. Chance accuracy is 25%.

GPT-5 Mini 75.0^*% Approx. 95% interval: 70.5–79.5% · 2025-10-30 · Older result

GPT-5 Nano 69.4^*% Approx. 95% interval: 64.0–74.9% · 2025-10-30 · Older result

Novel problem solving

Mystery Game Puzzles · 25%

Best-next-move puzzles in an undisclosed game variant. A narrow reasoning test, not general intelligence.

GPT-5 Mini 5.0^*% Approx. 95% interval: 0.7–9.3% · 2026-08-27

GPT-5 Nano 8.0^*% Approx. 95% interval: 2.7–13.3% · 2026-08-27

Advanced mathematics

FrontierMath Tiers 1-3 · 25%

Private v2 mathematics problems with Python tools. Earlier versions and Tier 4 are excluded.

GPT-5 Mini 46.7^*% Approx. 95% interval: 40.9–52.5% · 2026-06-12

GPT-5 Nano 20.0^*% Approx. 95% interval: 15.3–24.7% · 2026-06-12

Factual answers

SimpleQA Verified · 25%

Correct factual answers under a guessing prompt, graded by a model. Not a hallucination or abstention score.

GPT-5 Mini 21.6^*% Approx. 95% interval: 19.0–24.2% · 2026-08-10

GPT-5 Nano 11.7^*% Approx. 95% interval: 9.7–13.7% · 2026-08-10

Competition mathematics

OTIS Mock AIME 2024-2025 · Separate result

Competition problems with model-extracted answers checked against an exact key. Earlier grading procedures are excluded.

GPT-5 Mini 86.7^*% Approx. 95% interval: 78.3–95.1% · 2025-10-30 · Older result

GPT-5 Nano 81.1^*% Approx. 95% interval: 72.0–90.2% · 2025-10-31 · Older result

Research mathematics

FrontierMath Tier 4 v2 · Separate result

The hardest private FrontierMath problems, corrected v2 set, with Python tools. Separate from Tiers 1-3 in the core score.

GPT-5 Mini 12.2^*% Approx. 95% interval: 2.1–22.3% · 2026-06-12

GPT-5 Nano 2.4^*% Approx. 95% interval: 0.0–7.2% · 2026-06-12

Bars show reported task scores. Whiskers are approximate 95% intervals where standard errors are available. Different reasoning budgets are not equal-cost tests. Missing results are never zero.

Run it locally

Inference · memory estimate

GPT-5 Mini

Closed weights. No supported local deployment or GPU requirement is published.

GPT-5 Nano

Closed weights. No supported local deployment or GPU requirement is published.

GPT-5 Mini vs GPT-5 Nano
At a glance	Models GPT-5 Mini openai	Models GPT-5 Nano openai
Release date	2025-08-07	2025-08-07
Capabilities
Context window	400,000 tokens	400,000 tokens
Maximum input	272,000 tokens^*	272,000 tokens^*
Maximum output	128,000 tokens	128,000 tokens
Reasoning	Yes	Yes
Tool calling	Yes	Yes
Structured output	Yes^*	Yes^*
Open weights	No	No
Knowledge cutoff	2024-05-30	2024-05-30
API mode	Chat^*	Chat^*
Image input	Yes^*	Yes^*
PDF input	Yes^*	Yes^*
Parallel tool calls	Yes^*	Yes^*
Prompt caching	Yes^*	Yes^*
System messages	Yes^*	Yes^*
Tool choice	Yes^*	Yes^*
Web search	Yes^*	Yes^*
Model license	Proprietary	Proprietary
Detailed specifications	Specifications · 35 fields cache read input token cost 2.5e-8^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-mini^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* cache read input token cost batches 1.25e-8^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-mini^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* cache read input token cost flex 1.25e-8^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-mini^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* cache read input token cost priority 4.5e-8^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-mini^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* input cost per token 2.5e-7^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-mini^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* input cost per token batches 1.25e-7^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-mini^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* input cost per token flex 1.25e-7^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-mini^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* input cost per token priority 4.5e-7^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-mini^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* litellm provider openai^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-mini^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* max input tokens 272,000^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-mini^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* max output tokens 128,000^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-mini^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* max tokens 128,000^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-mini^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* mode chat^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-mini^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* output cost per token 2.0e-6^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-mini^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* output cost per token batches 1.0e-6^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-mini^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* output cost per token flex 1.0e-6^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-mini^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* output cost per token priority 3.6e-6^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-mini^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* search context cost per query search_context_size_high 0.01^* search_context_size_low 0.01^* search_context_size_medium 0.01^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-mini^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* supported endpoints /v1/chat/completions^* /v1/batch^* /v1/responses^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-mini^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* supported modalities text^* image^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-mini^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* supported output modalities text^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-mini^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* supports function calling Yes^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-mini^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* supports minimal reasoning effort Yes^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-mini^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* supports native streaming Yes^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-mini^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* supports none reasoning effort No^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-mini^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* supports parallel function calling Yes^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-mini^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* supports pdf input Yes^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-mini^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* supports prompt caching Yes^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-mini^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* supports reasoning Yes^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-mini^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* supports response schema Yes^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-mini^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* supports system messages Yes^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-mini^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* supports tool choice Yes^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-mini^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* supports vision Yes^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-mini^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* supports web search Yes^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-mini^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* supports xhigh reasoning effort No^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-mini^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^*	Specifications · 33 fields cache read input token cost 5.0e-9^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-nano^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* cache read input token cost batches 2.5e-9^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-nano^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* cache read input token cost flex 2.5e-9^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-nano^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* input cost per token 5.0e-8^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-nano^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* input cost per token batches 2.5e-8^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-nano^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* input cost per token flex 2.5e-8^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-nano^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* input cost per token priority 2.5e-6^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-nano^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* litellm provider openai^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-nano^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* max input tokens 272,000^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-nano^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* max output tokens 128,000^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-nano^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* max tokens 128,000^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-nano^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* mode chat^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-nano^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* output cost per token 4.0e-7^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-nano^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* output cost per token batches 2.0e-7^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-nano^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* output cost per token flex 2.0e-7^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-nano^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* search context cost per query search_context_size_high 0.01^* search_context_size_low 0.01^* search_context_size_medium 0.01^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-nano^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* supported endpoints /v1/chat/completions^* /v1/batch^* /v1/responses^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-nano^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* supported modalities text^* image^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-nano^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* supported output modalities text^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-nano^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* supports function calling Yes^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-nano^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* supports minimal reasoning effort Yes^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-nano^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* supports native streaming Yes^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-nano^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* supports none reasoning effort No^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-nano^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* supports parallel function calling Yes^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-nano^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* supports pdf input Yes^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-nano^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* supports prompt caching Yes^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-nano^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* supports reasoning Yes^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-nano^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* supports response schema Yes^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-nano^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* supports system messages Yes^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-nano^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* supports tool choice Yes^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-nano^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* supports vision Yes^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-nano^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* supports web search Yes^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-nano^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^* supports xhigh reasoning effort No^* Scope and conditions native_context provider openai^* mode chat^* full_native_model_id gpt-5-nano^* pricing_scope Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.^*
Input modalities	text image	text image
Output modalities	text	text

API cost

USD / month

Input tokens / monthOutput tokens / month

Standard text-token estimate. Excludes taxes, caching, batch discounts, reasoning and multimodal surcharges. Provider limits and regional availability may differ.

GPT-5 Mini

Provider offer

GPT-5 Nano

Provider offer

Recorded results

Only the same benchmark, version, workload, scenario and unit share a row. Missing results are not zero. System results include the full submitted hardware and software configuration.

Test version

Arena text preference / chinesehuman preference · 2026-10-28 · rating

GPT-5 Mini

1,385.71 rating

Confidence interval: 1,372.02 – 1,399.40

1949 votes

Oct 28, 2026

Category: chinese
Evaluated configuration: GPT-5 mini (High)
Evaluation type: human preference

GPT-5 Nano

1,355.88 rating

Confidence interval: 1,329.95 – 1,381.80

512 votes

Oct 28, 2026

Category: chinese
Evaluated configuration: GPT-5 nano (High)
Evaluation type: human preference

Arena text preference / codinghuman preference · 2026-10-28 · rating

GPT-5 Mini

1,405.70 rating

Confidence interval: 1,397.40 – 1,414.00

5514 votes

Oct 28, 2026

Category: coding
Evaluated configuration: GPT-5 mini (High)
Evaluation type: human preference

GPT-5 Nano

1,350.82 rating

Confidence interval: 1,336.23 – 1,365.40

1665 votes

Oct 28, 2026

Category: coding
Evaluated configuration: GPT-5 nano (High)
Evaluation type: human preference

Arena text preference / creative_writinghuman preference · 2026-10-28 · rating

GPT-5 Mini

1,324.98 rating

Confidence interval: 1,314.77 – 1,335.20

3521 votes

Oct 28, 2026

Category: creative_writing
Evaluated configuration: GPT-5 mini (High)
Evaluation type: human preference

GPT-5 Nano

1,249.26 rating

Confidence interval: 1,229.57 – 1,268.95

910 votes

Oct 28, 2026

Category: creative_writing
Evaluated configuration: GPT-5 nano (High)
Evaluation type: human preference

Arena text preference / englishhuman preference · 2026-10-28 · rating

GPT-5 Mini

1,378.97 rating

Confidence interval: 1,372.81 – 1,385.14

11466 votes

Oct 28, 2026

Category: english
Evaluated configuration: GPT-5 mini (High)
Evaluation type: human preference

GPT-5 Nano

1,321.55 rating

Confidence interval: 1,311.69 – 1,331.40

3581 votes

Oct 28, 2026

Category: english
Evaluated configuration: GPT-5 nano (High)
Evaluation type: human preference

Arena text preference / exclude_tieshuman preference · 2026-10-28 · rating

GPT-5 Mini

1,342.32 rating

Confidence interval: 1,336.24 – 1,348.40

19505 votes

Oct 28, 2026

Category: exclude_ties
Evaluated configuration: GPT-5 mini (High)
Evaluation type: human preference

GPT-5 Nano

1,268.10 rating

Confidence interval: 1,258.63 – 1,277.56

5887 votes

Oct 28, 2026

Category: exclude_ties
Evaluated configuration: GPT-5 nano (High)
Evaluation type: human preference

Arena text preference / experthuman preference · 2026-10-28 · rating

GPT-5 Mini

1,378.29 rating

Confidence interval: 1,361.06 – 1,395.52

1230 votes

Oct 28, 2026

Category: expert
Evaluated configuration: GPT-5 mini (High)
Evaluation type: human preference

GPT-5 Nano

1,321.18 rating

Confidence interval: 1,289.08 – 1,353.27

338 votes

Oct 28, 2026

Category: expert
Evaluated configuration: GPT-5 nano (High)
Evaluation type: human preference

Arena text preference / frenchhuman preference · 2026-10-28 · rating

GPT-5 Mini

1,386.57 rating

Confidence interval: 1,356.74 – 1,416.41

413 votes

Oct 28, 2026

Category: french
Evaluated configuration: GPT-5 mini (High)
Evaluation type: human preference

GPT-5 Nano

Not reported for this test

Arena text preference / germanhuman preference · 2026-10-28 · rating

GPT-5 Mini

1,365.74 rating

Confidence interval: 1,341.63 – 1,389.85

592 votes

Oct 28, 2026

Category: german
Evaluated configuration: GPT-5 mini (High)
Evaluation type: human preference

GPT-5 Nano

1,326.69 rating

Confidence interval: 1,285.66 – 1,367.72

206 votes

Oct 28, 2026

Category: german
Evaluated configuration: GPT-5 nano (High)
Evaluation type: human preference

Arena text preference / hard_promptshuman preference · 2026-10-28 · rating

GPT-5 Mini

1,379.88 rating

Confidence interval: 1,373.91 – 1,385.85

12855 votes

Oct 28, 2026

Category: hard_prompts
Evaluated configuration: GPT-5 mini (High)
Evaluation type: human preference

GPT-5 Nano

1,328.24 rating

Confidence interval: 1,318.56 – 1,337.92

3782 votes

Oct 28, 2026

Category: hard_prompts
Evaluated configuration: GPT-5 nano (High)
Evaluation type: human preference

Arena text preference / hard_prompts_englishhuman preference · 2026-10-28 · rating

GPT-5 Mini

1,383.19 rating

Confidence interval: 1,374.96 – 1,391.43

5597 votes

Oct 28, 2026

Category: hard_prompts_english
Evaluated configuration: GPT-5 mini (High)
Evaluation type: human preference

GPT-5 Nano

1,338.58 rating

Confidence interval: 1,324.19 – 1,352.98

1694 votes

Oct 28, 2026

Category: hard_prompts_english
Evaluated configuration: GPT-5 nano (High)
Evaluation type: human preference

Arena text preference / industry_business_and_management_and_financial_operationshuman preference · 2026-10-28 · rating

GPT-5 Mini

1,357.92 rating

Confidence interval: 1,349.09 – 1,366.74

4830 votes

Oct 28, 2026

Category: industry_business_and_management_and_financial_operations
Evaluated configuration: GPT-5 mini (High)
Evaluation type: human preference

GPT-5 Nano

1,310.77 rating

Confidence interval: 1,295.36 – 1,326.19

1521 votes

Oct 28, 2026

Category: industry_business_and_management_and_financial_operations
Evaluated configuration: GPT-5 nano (High)
Evaluation type: human preference

Arena text preference / industry_entertainment_and_sports_and_mediahuman preference · 2026-10-28 · rating

GPT-5 Mini

1,339.92 rating

Confidence interval: 1,331.11 – 1,348.72

4845 votes

Oct 28, 2026

Category: industry_entertainment_and_sports_and_media
Evaluated configuration: GPT-5 mini (High)
Evaluation type: human preference

GPT-5 Nano

1,268.63 rating

Confidence interval: 1,252.57 – 1,284.69

1358 votes

Oct 28, 2026

Category: industry_entertainment_and_sports_and_media
Evaluated configuration: GPT-5 nano (High)
Evaluation type: human preference

Arena text preference / industry_legal_and_governmenthuman preference · 2026-10-28 · rating

GPT-5 Mini

1,375.10 rating

Confidence interval: 1,360.76 – 1,389.45

1691 votes

Oct 28, 2026

Category: industry_legal_and_government
Evaluated configuration: GPT-5 mini (High)
Evaluation type: human preference

GPT-5 Nano

1,298.68 rating

Confidence interval: 1,271.71 – 1,325.64

470 votes

Oct 28, 2026

Category: industry_legal_and_government
Evaluated configuration: GPT-5 nano (High)
Evaluation type: human preference

Arena text preference / industry_life_and_physical_and_social_sciencehuman preference · 2026-10-28 · rating

GPT-5 Mini

1,370.58 rating

Confidence interval: 1,361.24 – 1,379.93

4208 votes

Oct 28, 2026

Category: industry_life_and_physical_and_social_science
Evaluated configuration: GPT-5 mini (High)
Evaluation type: human preference

GPT-5 Nano

1,319.03 rating

Confidence interval: 1,302.78 – 1,335.27

1309 votes

Oct 28, 2026

Category: industry_life_and_physical_and_social_science
Evaluated configuration: GPT-5 nano (High)
Evaluation type: human preference

Arena text preference / industry_mathematicalhuman preference · 2026-10-28 · rating

GPT-5 Mini

1,380.34 rating

Confidence interval: 1,363.95 – 1,396.73

1303 votes

Oct 28, 2026

Category: industry_mathematical
Evaluated configuration: GPT-5 mini (High)
Evaluation type: human preference

GPT-5 Nano

1,326.89 rating

Confidence interval: 1,298.54 – 1,355.24

439 votes

Oct 28, 2026

Category: industry_mathematical
Evaluated configuration: GPT-5 nano (High)
Evaluation type: human preference

Arena text preference / industry_medicine_and_healthcarehuman preference · 2026-10-28 · rating

GPT-5 Mini

1,379.54 rating

Confidence interval: 1,363.76 – 1,395.32

1436 votes

Oct 28, 2026

Category: industry_medicine_and_healthcare
Evaluated configuration: GPT-5 mini (High)
Evaluation type: human preference

GPT-5 Nano

1,335.07 rating

Confidence interval: 1,308.11 – 1,362.02

470 votes

Oct 28, 2026

Category: industry_medicine_and_healthcare
Evaluated configuration: GPT-5 nano (High)
Evaluation type: human preference

Arena text preference / industry_software_and_it_serviceshuman preference · 2026-10-28 · rating

GPT-5 Mini

1,393.69 rating

Confidence interval: 1,386.89 – 1,400.49

8914 votes

Oct 28, 2026

Category: industry_software_and_it_services
Evaluated configuration: GPT-5 mini (High)
Evaluation type: human preference

GPT-5 Nano

1,344.91 rating

Confidence interval: 1,333.74 – 1,356.08

2810 votes

Oct 28, 2026

Category: industry_software_and_it_services
Evaluated configuration: GPT-5 nano (High)
Evaluation type: human preference

Arena text preference / industry_writing_and_literature_and_languagehuman preference · 2026-10-28 · rating

GPT-5 Mini

1,344.07 rating

Confidence interval: 1,336.20 – 1,351.95

6141 votes

Oct 28, 2026

Category: industry_writing_and_literature_and_language
Evaluated configuration: GPT-5 mini (High)
Evaluation type: human preference

GPT-5 Nano

1,283.11 rating

Confidence interval: 1,269.01 – 1,297.22

1726 votes

Oct 28, 2026

Category: industry_writing_and_literature_and_language
Evaluated configuration: GPT-5 nano (High)
Evaluation type: human preference

Arena text preference / instruction_followinghuman preference · 2026-10-28 · rating

GPT-5 Mini

1,356.46 rating

Confidence interval: 1,349.00 – 1,363.92

7010 votes

Oct 28, 2026

Category: instruction_following
Evaluated configuration: GPT-5 mini (High)
Evaluation type: human preference

GPT-5 Nano

1,306.04 rating

Confidence interval: 1,292.96 – 1,319.12

2023 votes

Oct 28, 2026

Category: instruction_following
Evaluated configuration: GPT-5 nano (High)
Evaluation type: human preference

Arena text preference / japanesehuman preference · 2026-10-28 · rating

GPT-5 Mini

1,341.21 rating

Confidence interval: 1,312.76 – 1,369.66

460 votes

Oct 28, 2026

Category: japanese
Evaluated configuration: GPT-5 mini (High)
Evaluation type: human preference

GPT-5 Nano

1,226.65 rating

Confidence interval: 1,177.17 – 1,276.13

147 votes

Oct 28, 2026

Category: japanese
Evaluated configuration: GPT-5 nano (High)
Evaluation type: human preference

Arena text preference / koreanhuman preference · 2026-10-28 · rating

GPT-5 Mini

1,306.53 rating

Confidence interval: 1,282.34 – 1,330.71

663 votes

Oct 28, 2026

Category: korean
Evaluated configuration: GPT-5 mini (High)
Evaluation type: human preference

GPT-5 Nano

1,269.04 rating

Confidence interval: 1,228.46 – 1,309.63

225 votes

Oct 28, 2026

Category: korean
Evaluated configuration: GPT-5 nano (High)
Evaluation type: human preference

Arena text preference / longer_queryhuman preference · 2026-10-28 · rating

GPT-5 Mini

1,354.58 rating

Confidence interval: 1,346.54 – 1,362.63

5903 votes

Oct 28, 2026

Category: longer_query
Evaluated configuration: GPT-5 mini (High)
Evaluation type: human preference

GPT-5 Nano

1,311.81 rating

Confidence interval: 1,297.43 – 1,326.20

1631 votes

Oct 28, 2026

Category: longer_query
Evaluated configuration: GPT-5 nano (High)
Evaluation type: human preference

Arena text preference / mathhuman preference · 2026-10-28 · rating

GPT-5 Mini

1,378.02 rating

Confidence interval: 1,362.64 – 1,393.40

1456 votes

Oct 28, 2026

Category: math
Evaluated configuration: GPT-5 mini (High)
Evaluation type: human preference

GPT-5 Nano

1,318.26 rating

Confidence interval: 1,291.78 – 1,344.75

493 votes

Oct 28, 2026

Category: math
Evaluated configuration: GPT-5 nano (High)
Evaluation type: human preference

Arena text preference / multi_turnhuman preference · 2026-10-28 · rating

GPT-5 Mini

1,362.84 rating

Confidence interval: 1,353.64 – 1,372.03

4394 votes

Oct 28, 2026

Category: multi_turn
Evaluated configuration: GPT-5 mini (High)
Evaluation type: human preference

GPT-5 Nano

1,311.24 rating

Confidence interval: 1,295.58 – 1,326.90

1420 votes

Oct 28, 2026

Category: multi_turn
Evaluated configuration: GPT-5 nano (High)
Evaluation type: human preference

Arena text preference / non_englishhuman preference · 2026-10-28 · rating

GPT-5 Mini

1,362.39 rating

Confidence interval: 1,356.76 – 1,368.03

15286 votes

Oct 28, 2026

Category: non_english
Evaluated configuration: GPT-5 mini (High)
Evaluation type: human preference

GPT-5 Nano

1,312.63 rating

Confidence interval: 1,303.84 – 1,321.42

4551 votes

Oct 28, 2026

Category: non_english
Evaluated configuration: GPT-5 nano (High)
Evaluation type: human preference

Arena text preference / overallhuman preference · 2026-10-28 · rating

GPT-5 Mini

1,373.21 rating

Confidence interval: 1,368.65 – 1,377.76

26793 votes

Oct 28, 2026

Category: overall
Evaluated configuration: GPT-5 mini (High)
Evaluation type: human preference

GPT-5 Nano

1,319.89 rating

Confidence interval: 1,313.05 – 1,326.74

8134 votes

Oct 28, 2026

Category: overall
Evaluated configuration: GPT-5 nano (High)
Evaluation type: human preference

Arena text preference / polishhuman preference · 2026-10-28 · rating

GPT-5 Mini

1,378.12 rating

Confidence interval: 1,362.66 – 1,393.58

1562 votes

Oct 28, 2026

Category: polish
Evaluated configuration: GPT-5 mini (High)
Evaluation type: human preference

GPT-5 Nano

1,313.14 rating

Confidence interval: 1,288.20 – 1,338.09

572 votes

Oct 28, 2026

Category: polish
Evaluated configuration: GPT-5 nano (High)
Evaluation type: human preference

Arena text preference / russianhuman preference · 2026-10-28 · rating

GPT-5 Mini

1,361.46 rating

Confidence interval: 1,348.37 – 1,374.55

1870 votes

Oct 28, 2026

Category: russian
Evaluated configuration: GPT-5 mini (High)
Evaluation type: human preference

GPT-5 Nano

1,296.28 rating

Confidence interval: 1,272.70 – 1,319.86

601 votes

Oct 28, 2026

Category: russian
Evaluated configuration: GPT-5 nano (High)
Evaluation type: human preference

Arena text preference / spanishhuman preference · 2026-10-28 · rating

GPT-5 Mini

1,354.58 rating

Confidence interval: 1,333.71 – 1,375.44

856 votes

Oct 28, 2026

Category: spanish
Evaluated configuration: GPT-5 mini (High)
Evaluation type: human preference

GPT-5 Nano

1,360.34 rating

Confidence interval: 1,321.17 – 1,399.52

237 votes

Oct 28, 2026

Category: spanish
Evaluated configuration: GPT-5 nano (High)
Evaluation type: human preference

Arena text preference / chinesehuman preference · 2026-10-15 · rating

GPT-5 Mini

1,385.91 rating

Confidence interval: 1,372.22 – 1,399.60

1951 votes

Oct 15, 2026

Category: chinese
Evaluated configuration: gpt-5-mini-high
Evaluation type: human preference

GPT-5 Nano

1,356.01 rating

Confidence interval: 1,330.08 – 1,381.93

512 votes

Oct 15, 2026

Category: chinese
Evaluated configuration: gpt-5-nano-high
Evaluation type: human preference

Arena text preference / codinghuman preference · 2026-10-15 · rating

GPT-5 Mini

1,405.78 rating

Confidence interval: 1,397.48 – 1,414.08

5520 votes

Oct 15, 2026

Category: coding
Evaluated configuration: gpt-5-mini-high
Evaluation type: human preference

GPT-5 Nano

1,351.14 rating

Confidence interval: 1,336.55 – 1,365.72

1664 votes

Oct 15, 2026

Category: coding
Evaluated configuration: gpt-5-nano-high
Evaluation type: human preference

Arena text preference / creative_writinghuman preference · 2026-10-15 · rating

GPT-5 Mini

1,325.29 rating

Confidence interval: 1,315.07 – 1,335.51

3520 votes

Oct 15, 2026

Category: creative_writing
Evaluated configuration: gpt-5-mini-high
Evaluation type: human preference

GPT-5 Nano

1,249.44 rating

Confidence interval: 1,229.77 – 1,269.12

911 votes

Oct 15, 2026

Category: creative_writing
Evaluated configuration: gpt-5-nano-high
Evaluation type: human preference

Arena text preference / englishhuman preference · 2026-10-15 · rating

GPT-5 Mini

1,378.74 rating

Confidence interval: 1,372.57 – 1,384.90

11472 votes

Oct 15, 2026

Category: english
Evaluated configuration: gpt-5-mini-high
Evaluation type: human preference

GPT-5 Nano

1,321.53 rating

Confidence interval: 1,311.67 – 1,331.38

3580 votes

Oct 15, 2026

Category: english
Evaluated configuration: gpt-5-nano-high
Evaluation type: human preference

Arena text preference / exclude_tieshuman preference · 2026-10-15 · rating

GPT-5 Mini

1,342.42 rating

Confidence interval: 1,336.33 – 1,348.50

19513 votes

Oct 15, 2026

Category: exclude_ties
Evaluated configuration: gpt-5-mini-high
Evaluation type: human preference

GPT-5 Nano

1,268.19 rating

Confidence interval: 1,258.73 – 1,277.65

5885 votes

Oct 15, 2026

Category: exclude_ties
Evaluated configuration: gpt-5-nano-high
Evaluation type: human preference

Arena text preference / experthuman preference · 2026-10-15 · rating

GPT-5 Mini

1,378.57 rating

Confidence interval: 1,361.34 – 1,395.81

1229 votes

Oct 15, 2026

Category: expert
Evaluated configuration: gpt-5-mini-high
Evaluation type: human preference

GPT-5 Nano

1,321.22 rating

Confidence interval: 1,289.12 – 1,353.32

338 votes

Oct 15, 2026

Category: expert
Evaluated configuration: gpt-5-nano-high
Evaluation type: human preference

Arena text preference / frenchhuman preference · 2026-10-15 · rating

GPT-5 Mini

1,386.06 rating

Confidence interval: 1,356.23 – 1,415.90

413 votes

Oct 15, 2026

Category: french
Evaluated configuration: gpt-5-mini-high
Evaluation type: human preference

GPT-5 Nano

Not reported for this test

Arena text preference / germanhuman preference · 2026-10-15 · rating

GPT-5 Mini

1,365.70 rating

Confidence interval: 1,341.59 – 1,389.81

592 votes

Oct 15, 2026

Category: german
Evaluated configuration: gpt-5-mini-high
Evaluation type: human preference

GPT-5 Nano

1,326.62 rating

Confidence interval: 1,285.59 – 1,367.64

206 votes

Oct 15, 2026

Category: german
Evaluated configuration: gpt-5-nano-high
Evaluation type: human preference

Arena text preference / hard_promptshuman preference · 2026-10-15 · rating

GPT-5 Mini

1,379.96 rating

Confidence interval: 1,373.98 – 1,385.93

12861 votes

Oct 15, 2026

Category: hard_prompts
Evaluated configuration: gpt-5-mini-high
Evaluation type: human preference

GPT-5 Nano

1,328.43 rating

Confidence interval: 1,318.74 – 1,338.12

3779 votes

Oct 15, 2026

Category: hard_prompts
Evaluated configuration: gpt-5-nano-high
Evaluation type: human preference

Arena text preference / hard_prompts_englishhuman preference · 2026-10-15 · rating

GPT-5 Mini

1,382.79 rating

Confidence interval: 1,374.56 – 1,391.02

5602 votes

Oct 15, 2026

Category: hard_prompts_english
Evaluated configuration: gpt-5-mini-high
Evaluation type: human preference

GPT-5 Nano

1,338.50 rating

Confidence interval: 1,324.11 – 1,352.90

1694 votes

Oct 15, 2026

Category: hard_prompts_english
Evaluated configuration: gpt-5-nano-high
Evaluation type: human preference

Arena text preference / industry_business_and_management_and_financial_operationshuman preference · 2026-10-15 · rating

GPT-5 Mini

1,357.83 rating

Confidence interval: 1,349.01 – 1,366.66

4833 votes

Oct 15, 2026

Category: industry_business_and_management_and_financial_operations
Evaluated configuration: gpt-5-mini-high
Evaluation type: human preference

GPT-5 Nano

1,310.92 rating

Confidence interval: 1,295.51 – 1,326.33

1521 votes

Oct 15, 2026

Category: industry_business_and_management_and_financial_operations
Evaluated configuration: gpt-5-nano-high
Evaluation type: human preference

Arena text preference / industry_entertainment_and_sports_and_mediahuman preference · 2026-10-15 · rating

GPT-5 Mini

1,339.94 rating

Confidence interval: 1,331.14 – 1,348.75

4844 votes

Oct 15, 2026

Category: industry_entertainment_and_sports_and_media
Evaluated configuration: gpt-5-mini-high
Evaluation type: human preference

GPT-5 Nano

1,268.57 rating

Confidence interval: 1,252.51 – 1,284.62

1357 votes

Oct 15, 2026

Category: industry_entertainment_and_sports_and_media
Evaluated configuration: gpt-5-nano-high
Evaluation type: human preference

Arena text preference / industry_legal_and_governmenthuman preference · 2026-10-15 · rating

GPT-5 Mini

1,374.75 rating

Confidence interval: 1,360.41 – 1,389.08

1693 votes

Oct 15, 2026

Category: industry_legal_and_government
Evaluated configuration: gpt-5-mini-high
Evaluation type: human preference

GPT-5 Nano

1,297.40 rating

Confidence interval: 1,270.42 – 1,324.38

469 votes

Oct 15, 2026

Category: industry_legal_and_government
Evaluated configuration: gpt-5-nano-high
Evaluation type: human preference

Arena text preference / industry_life_and_physical_and_social_sciencehuman preference · 2026-10-15 · rating

GPT-5 Mini

1,370.68 rating

Confidence interval: 1,361.33 – 1,380.02

4208 votes

Oct 15, 2026

Category: industry_life_and_physical_and_social_science
Evaluated configuration: gpt-5-mini-high
Evaluation type: human preference

GPT-5 Nano

1,318.66 rating

Confidence interval: 1,302.40 – 1,334.92

1307 votes

Oct 15, 2026

Category: industry_life_and_physical_and_social_science
Evaluated configuration: gpt-5-nano-high
Evaluation type: human preference

Arena text preference / industry_mathematicalhuman preference · 2026-10-15 · rating

GPT-5 Mini

1,380.13 rating

Confidence interval: 1,363.75 – 1,396.52

1304 votes

Oct 15, 2026

Category: industry_mathematical
Evaluated configuration: gpt-5-mini-high
Evaluation type: human preference

GPT-5 Nano

1,325.99 rating

Confidence interval: 1,297.62 – 1,354.36

440 votes

Oct 15, 2026

Category: industry_mathematical
Evaluated configuration: gpt-5-nano-high
Evaluation type: human preference

Arena text preference / industry_medicine_and_healthcarehuman preference · 2026-10-15 · rating

GPT-5 Mini

1,379.61 rating

Confidence interval: 1,363.84 – 1,395.39

1436 votes

Oct 15, 2026

Category: industry_medicine_and_healthcare
Evaluated configuration: gpt-5-mini-high
Evaluation type: human preference

GPT-5 Nano

1,335.68 rating

Confidence interval: 1,308.70 – 1,362.66

470 votes

Oct 15, 2026

Category: industry_medicine_and_healthcare
Evaluated configuration: gpt-5-nano-high
Evaluation type: human preference

Arena text preference / industry_software_and_it_serviceshuman preference · 2026-10-15 · rating

GPT-5 Mini

1,393.73 rating

Confidence interval: 1,386.93 – 1,400.53

8921 votes

Oct 15, 2026

Category: industry_software_and_it_services
Evaluated configuration: gpt-5-mini-high
Evaluation type: human preference

GPT-5 Nano

1,345.12 rating

Confidence interval: 1,333.94 – 1,356.29

2809 votes

Oct 15, 2026

Category: industry_software_and_it_services
Evaluated configuration: gpt-5-nano-high
Evaluation type: human preference

Arena text preference / industry_writing_and_literature_and_languagehuman preference · 2026-10-15 · rating

GPT-5 Mini

1,344.23 rating

Confidence interval: 1,336.35 – 1,352.10

6141 votes

Oct 15, 2026

Category: industry_writing_and_literature_and_language
Evaluated configuration: gpt-5-mini-high
Evaluation type: human preference

GPT-5 Nano

1,283.25 rating

Confidence interval: 1,269.15 – 1,297.34

1727 votes

Oct 15, 2026

Category: industry_writing_and_literature_and_language
Evaluated configuration: gpt-5-nano-high
Evaluation type: human preference

Arena text preference / instruction_followinghuman preference · 2026-10-15 · rating

GPT-5 Mini

1,356.78 rating

Confidence interval: 1,349.32 – 1,364.24

7012 votes

Oct 15, 2026

Category: instruction_following
Evaluated configuration: gpt-5-mini-high
Evaluation type: human preference

GPT-5 Nano

1,306.05 rating

Confidence interval: 1,292.98 – 1,319.13

2023 votes

Oct 15, 2026

Category: instruction_following
Evaluated configuration: gpt-5-nano-high
Evaluation type: human preference

Arena text preference / japanesehuman preference · 2026-10-15 · rating

GPT-5 Mini

1,340.91 rating

Confidence interval: 1,312.41 – 1,369.40

459 votes

Oct 15, 2026

Category: japanese
Evaluated configuration: gpt-5-mini-high
Evaluation type: human preference

GPT-5 Nano

1,226.40 rating

Confidence interval: 1,176.92 – 1,275.89

147 votes

Oct 15, 2026

Category: japanese
Evaluated configuration: gpt-5-nano-high
Evaluation type: human preference

Arena text preference / koreanhuman preference · 2026-10-15 · rating

GPT-5 Mini

1,307.06 rating

Confidence interval: 1,282.86 – 1,331.26

662 votes

Oct 15, 2026

Category: korean
Evaluated configuration: gpt-5-mini-high
Evaluation type: human preference

GPT-5 Nano

1,269.29 rating

Confidence interval: 1,228.70 – 1,309.88

225 votes

Oct 15, 2026

Category: korean
Evaluated configuration: gpt-5-nano-high
Evaluation type: human preference

Arena text preference / longer_queryhuman preference · 2026-10-15 · rating

GPT-5 Mini

1,354.84 rating

Confidence interval: 1,346.80 – 1,362.89

5902 votes

Oct 15, 2026

Category: longer_query
Evaluated configuration: gpt-5-mini-high
Evaluation type: human preference

GPT-5 Nano

1,312.33 rating

Confidence interval: 1,297.93 – 1,326.72

1628 votes

Oct 15, 2026

Category: longer_query
Evaluated configuration: gpt-5-nano-high
Evaluation type: human preference

Arena text preference / mathhuman preference · 2026-10-15 · rating

GPT-5 Mini

1,377.40 rating

Confidence interval: 1,362.02 – 1,392.79

1457 votes

Oct 15, 2026

Category: math
Evaluated configuration: gpt-5-mini-high
Evaluation type: human preference

GPT-5 Nano

1,317.44 rating

Confidence interval: 1,290.93 – 1,343.95

494 votes

Oct 15, 2026

Category: math
Evaluated configuration: gpt-5-nano-high
Evaluation type: human preference

Arena text preference / multi_turnhuman preference · 2026-10-15 · rating

GPT-5 Mini

1,362.87 rating

Confidence interval: 1,353.68 – 1,372.07

4396 votes

Oct 15, 2026

Category: multi_turn
Evaluated configuration: gpt-5-mini-high
Evaluation type: human preference

GPT-5 Nano

1,311.36 rating

Confidence interval: 1,295.69 – 1,327.03

1418 votes

Oct 15, 2026

Category: multi_turn
Evaluated configuration: gpt-5-nano-high
Evaluation type: human preference

Arena text preference / non_englishhuman preference · 2026-10-15 · rating

GPT-5 Mini

1,362.69 rating

Confidence interval: 1,357.05 – 1,368.33

15291 votes

Oct 15, 2026

Category: non_english
Evaluated configuration: gpt-5-mini-high
Evaluation type: human preference

GPT-5 Nano

1,312.80 rating

Confidence interval: 1,304.01 – 1,321.59

4550 votes

Oct 15, 2026

Category: non_english
Evaluated configuration: gpt-5-nano-high
Evaluation type: human preference

Arena text preference / overallhuman preference · 2026-10-15 · rating

GPT-5 Mini

1,373.28 rating

Confidence interval: 1,368.73 – 1,377.84

26804 votes

Oct 15, 2026

Category: overall
Evaluated configuration: gpt-5-mini-high
Evaluation type: human preference

GPT-5 Nano

1,319.94 rating

Confidence interval: 1,313.09 – 1,326.78

8132 votes

Oct 15, 2026

Category: overall
Evaluated configuration: gpt-5-nano-high
Evaluation type: human preference

Arena text preference / polishhuman preference · 2026-10-15 · rating

GPT-5 Mini

1,378.39 rating

Confidence interval: 1,362.93 – 1,393.84

1563 votes

Oct 15, 2026

Category: polish
Evaluated configuration: gpt-5-mini-high
Evaluation type: human preference

GPT-5 Nano

1,313.75 rating

Confidence interval: 1,288.79 – 1,338.70

571 votes

Oct 15, 2026

Category: polish
Evaluated configuration: gpt-5-nano-high
Evaluation type: human preference

Arena text preference / russianhuman preference · 2026-10-15 · rating

GPT-5 Mini

1,361.77 rating

Confidence interval: 1,348.68 – 1,374.85

1872 votes

Oct 15, 2026

Category: russian
Evaluated configuration: gpt-5-mini-high
Evaluation type: human preference

GPT-5 Nano

1,295.89 rating

Confidence interval: 1,272.36 – 1,319.43

603 votes

Oct 15, 2026

Category: russian
Evaluated configuration: gpt-5-nano-high
Evaluation type: human preference

Arena text preference / spanishhuman preference · 2026-10-15 · rating

GPT-5 Mini

1,354.75 rating

Confidence interval: 1,333.87 – 1,375.64

855 votes

Oct 15, 2026

Category: spanish
Evaluated configuration: gpt-5-mini-high
Evaluation type: human preference

GPT-5 Nano

1,360.29 rating

Confidence interval: 1,321.11 – 1,399.47

237 votes

Oct 15, 2026

Category: spanish
Evaluated configuration: gpt-5-nano-high
Evaluation type: human preference

* Provisional data