supermarket

Vision-language

GPT-5 nano

By OpenAI, in the GPT-5 family.

Released
7 Aug 2025
Status
Deprecated
Retired
11 Dec 2026
Context
400K
Max output
128K
Input per 1M tokens
$0.05
Output per 1M tokens
$0.40
Open weights
No
License
proprietary
Parameters
—
Knowledge cutoff
31 May 2024
Reasoning
Yes
Modalities
text, image → text

Category scores

Out of 100, from independent and third-party results only, by method category-2026-09-29.1.

Category scores
CategoryScoreRankCoverage
Knowledge & reasoning16.3842 of 4
Vision & multimodal understanding5.8Not ranked1 of 3
Factuality & hallucination11.7661 of 1
Human preference30.0Not ranked2 of 5

Benchmark results

Every result we hold for this model, with who ran it and where it was published.

Knowledge & reasoning

Knowledge & reasoning results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
Chess PuzzlesPuzzles solved27.0%Independenthigh7 Aug 2026Epoch AI
Chess PuzzlesPuzzles solved15.0%Independentlow13 Jul 2026Epoch AI
Chess PuzzlesPuzzles solved1.0%Independentminimal13 Jul 2026Epoch AI
DTBench (Decision Theory Benchmark)Accuracy (capability questions)62.7%Third partyhigh—https://conceptualreasoning.ai/dtbench via Epoch AI
LMCA (Language Model Conceptual Argumentation)Correlation with expert ratings (Pearson r)0.0793Third partyhigh—https://conceptualreasoning.ai/lmca via Epoch AI
Mystery Game PuzzlesPuzzles solved9.0%Independentmedium27 Aug 2026Epoch AI
Mystery Game PuzzlesPuzzles solved9.0%Independentlow27 Aug 2026Epoch AI
Mystery Game PuzzlesPuzzles solved8.0%Independenthigh27 Aug 2026Epoch AI
Mystery Game PuzzlesPuzzles solved5.0%Independentminimal27 Aug 2026Epoch AI

Scientific reasoning

Scientific reasoning results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
GPQA DiamondAccuracy69.4%Independenthigh30 Oct 2025Epoch AI
GPQA DiamondAccuracy67.4%Independentmedium7 Aug 2025Epoch AI
GPQA DiamondAccuracy57.6%Independentlow13 Jul 2026Epoch AI
GPQA DiamondAccuracy48.5%Independentminimal13 Jul 2026Epoch AI

Mathematics

Mathematics results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
FrontierMath Tier 4Accuracy2.4%Independenthigh12 Jun 2026Epoch AI
FrontierMath Tier 4 (v1)Accuracy2.1%Independentmedium7 Aug 2025Epoch AI
FrontierMath Tier 4 (v1)Accuracy0.0%Independenthigh30 Oct 2025Epoch AI
FrontierMath Tiers 1-3Accuracy20.0%Independenthigh12 Jun 2026Epoch AI
FrontierMath Tiers 1-3Accuracy6.0%Independentlow27 Aug 2026Epoch AI
FrontierMath Tiers 1-3Accuracy1.8%Independentminimal27 Aug 2026Epoch AI
FrontierMath Tiers 1-3 (v1)Accuracy8.3%Independenthigh30 Oct 2025Epoch AI
FrontierMath Tiers 1-3 (v1)Accuracy7.2%Independentmedium7 Aug 2025Epoch AI
MATH Level 5Accuracy95.2%Independentmedium20 Aug 2025Epoch AI
MATH Level 5Accuracy94.9%Independenthigh20 Aug 2025Epoch AI
OTIS Mock AIME 2024-2025Accuracy81.1%Independenthigh31 Oct 2025Epoch AI
OTIS Mock AIME 2024-2025Accuracy74.2%Independentmedium7 Aug 2025Epoch AI
OTIS Mock AIME 2024-2025Accuracy46.7%Independentlow13 Jul 2026Epoch AI
OTIS Mock AIME 2024-2025Accuracy35.6%Independentminimal13 Jul 2026Epoch AI

Coding & software engineering

Coding & software engineering results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
Terminal-Bench 2.0Accuracy (mean over trials)21.8%Third partyunknown—https://www.tbench.ai/leaderboard/terminal-bench/2.0 via Epoch AI
Terminal-Bench 2.0Accuracy (mean over trials)11.5%Third partymedium—Terminal-Bench v2 Leaderboard via Epoch AI
Terminal-Bench 2.0Accuracy (mean over trials)11.5%Third partyunknown—https://www.tbench.ai/leaderboard/terminal-bench/2.0 via Epoch AI
Terminal-Bench 2.0Accuracy (mean over trials)9.9%Third partymedium—Terminal-Bench v2 Leaderboard via Epoch AI
Terminal-Bench 2.0Accuracy (mean over trials)9.9%Third partyunknown—https://www.tbench.ai/leaderboard/terminal-bench/2.0 via Epoch AI
Terminal-Bench 2.0Accuracy (mean over trials)7.9%Third partymedium—Terminal-Bench v2 Leaderboard via Epoch AI
Terminal-Bench 2.0Accuracy (mean over trials)7.9%Third partyunknown—https://www.tbench.ai/leaderboard/terminal-bench/2.0 via Epoch AI
Terminal-Bench 2.0Accuracy (mean over trials)7.0%Third partymedium—Terminal-Bench v2 Leaderboard via Epoch AI
Terminal-Bench 2.0Accuracy (mean over trials)7.0%Third partyunknown—https://www.tbench.ai/leaderboard/terminal-bench/2.0 via Epoch AI
WeirdML (v2)Average accuracy38.1%Third partyhigh—https://htihle.github.io/weirdml.html via Epoch AI
WeirdML (v2)Average accuracy25.9%Third partylow—https://htihle.github.io/weirdml.html via Epoch AI

Long context

Long context results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
Fiction.LiveBench% correct at 16K tokens (Epoch's column)44.4%Third partymedium—Fiction.live leaderboard via Epoch AI

Vision & multimodal understanding

Vision & multimodal understanding results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
Visual Physics Comprehension Test (VPCT)Correct37.2%Third partyhigh—VPCT leaderboard via Epoch AI
Visual Physics Comprehension Test (VPCT)Correct35.4%Third partymedium—VPCT leaderboard via Epoch AI

Factuality & hallucination

Factuality & hallucination results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
SimpleQA VerifiedAccuracy (% correct)11.7%Independenthigh10 Aug 2026Epoch AI

Human preference

Human preference results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
Text Arena (arena.ai)Arena score (Elo scale)1338Independenthigh25 Sep 2026Arena
Text Arena (arena.ai)Votes8111Independenthigh25 Sep 2026Arena
Vision Arena (arena.ai)Arena score (Elo scale)1146Independenthigh28 Sep 2026Arena
Vision Arena (arena.ai)Votes3934Independenthigh28 Sep 2026Arena

Other GPT-5 versions

Sources

GPT-5 nano: benchmarks, price and specs · supermarket.dev