GPT-5 nano

GPT-5 nano is a reasoning model from OpenAI in the GPT-5 family. 19 benchmarks count toward its score, in 6 categories.

availableShows if the model has enough results for an index.
IndexOverall score. 50 is the middle.43.2 ±6.2
CoverageShare of the index weight with results.75%
SpeedOutput tokens per second.85/s
Input / 1MUS dollars per 1M input tokens.$0.05 batch $0.025
Output / 1MUS dollars per 1M output tokens.$0.4 batch $0.2 US dollars per 1M output tokens in a batch.
ContextMaximum tokens in one request.400K
EloLMArena rating and rank.1320 (#218)

50 is the middle of the board. The range shows the doubt in the index. Batch work costs less.

8,128 votes. Elo shows what people prefer. It does not change the score.

CapabilitiesScore per category. 50 is the middle.

50 is the middle
AgenticMulti-step tasks with tools.
N/A
CodingCode writing and repair.
36.8
ReasoningLogic problems and puzzles.
44.2
MultimodalTasks with images and text.
42.8
KnowledgeFacts and expert knowledge.
37.4
MultilingualTasks in many languages.
68.8
InstructionTasks with strict rules in the prompt.
N/A
MathMath problems.
46.8

Results

19 counted
BenchmarkThe test name.CategoryThe capability that the test measures.ResultThe score from the publisher.IndexThis result on the index scale.RunThe settings of the run.DateDate of the result.Published byThe source of the result.
MATH level 5Math94.9%45.3high effortEpoch AI
MATH 500Math93.8%46.69 Jan 2026Vals AI
MGSMMultilingual89.3%9 Jan 2026Vals AI
AIMEMath81.2%51.616 Apr 2026Vals AI
OTIS Mock AIME 2024-2025Math81.1%56.9high effortEpoch AI
MMLU ProKnowledge76.1%40.31 Sept 2026Vals AI
MMMU ProMultimodal70.9%42.81 Sept 2026Vals AI
LiveCodeBenchCoding70.2%48.21 Sept 2026Vals AI
GPQA diamondKnowledge69.4%42.4high effortEpoch AI
GPQA DiamondKnowledge63.4%36.81 Sept 2026Vals AI
EuroEval SwedishMultilingual62.3%80.1EuroEval
EuroEval FrenchMultilingual53.9%69.6EuroEval
EuroEval ItalianMultilingual53.8%69.4EuroEval
EuroEval DutchMultilingual53.7%69.4EuroEval
EuroEval PolishMultilingual52.8%68.2EuroEval
EuroEval PortugueseMultilingual49.0%63.5EuroEval
EuroEval SpanishMultilingual49.0%63.5EuroEval
EuroEval GermanMultilingual47.0%61.0EuroEval
SWE-bench VerifiedCoding34.8%25.4medium effort · mini-SWE-agent26 Feb 2026SWE-bench team
SWE-bench VerifiedCoding34.8%25.4medium effort · mini-SWE-agent1 Sept 2026SWE-bench team
Chess PuzzlesReasoning27.0%58.4high effortEpoch AI
FrontierMath-Tiers-1-3-v2-PrivateMath20.0%42.7high effortEpoch AI
ARC-AGI-1 (semi-private)Reasoning16.7%35.1high effortARC Prize Foundation
SimpleQA VerifiedKnowledge11.7%32.4high effortEpoch AI
FrontierMath-2025-02-28-PrivateMath8.3%38.4high effortEpoch AI
Mystery Game PuzzlesReasoning8.0%42.4high effortEpoch AI
ARC-AGI-2 (semi-private)Reasoning2.6%41.0high effortARC Prize Foundation
FrontierMath-Tier-4-v2-PrivateMath2.4%49.1high effortEpoch AI
FrontierMath-Tier-4-2025-07-01-PrivateMath0.0%43.4high effortEpoch AI

19 benchmarks count, from 28 of 29 results. A grey row does not count. Too few models took that benchmark.

Sources

BenchLM benchmark aggregationUsed with attribution; per-benchmark results credited to their original authorsOpenRouter, collected directlyNo licence statedEpoch AI, collected directlyCC BY — free to use and redistribute with attributionVals AI, collected directlyNo licence stated. Read from the public leaderboard and credited to Vals AIEuroEval, collected directlyMIT — the leaderboard site and its CSV routes are in the licensed repositorySWE-bench team, collected directlyNo licence stated for the leaderboard. The harness repo is MITARC Prize Foundation, collected directlyNo licence stated. Their terms ask for written permission before commercial use

Same level, lower price

Qwen3.8-Flash-Next66.4 · Freedots3-note Preview65.5 · FreeApodex 1.1 Mini59.1 · Free

More from OpenAI

GPT-6 Astra83.5GPT-5.6 Sol76.7GPT-5.6 Terra73.4GPT-6 Sol76.9GPT-5.572.3