Inference360
OpenAI

o4-mini (high)

o4-mini (high) is a reasoning model from OpenAI in the o4-mini family. 20 benchmarks count toward its score, in 6 categories.

availableShows if the model has enough results for an index.
IndexOverall score out of 100.47.4 ±4.6
CoverageShare of the index weight with results.90%
SpeedOutput tokens per second.109/s
Input / 1MUS dollars per 1M input tokens.$1.1
Output / 1MUS dollars per 1M output tokens.$4.4
ContextMaximum tokens in one request.200K
EloLMArena rating and rank.1353 (#190)

The index is a score out of 100. The ± range shows how much it can change.

44,639 votes. Elo shows what people prefer. It does not change the score.

CapabilitiesScore per category, out of 100.

Out of 100
AgenticMulti-step tasks with tools.
36.2
CodingCode writing and repair.
46.4
ReasoningLogic problems and puzzles.
48.5
MultimodalTasks with images and text.
57.0
KnowledgeFacts and expert knowledge.
45.3
MultilingualTasks in many languages.
N/A
InstructionTasks with strict rules in the prompt.
N/A
MathMath problems.
51.2

Results

20 counted
BenchmarkThe test name.CategoryThe capability that the test measures.ResultThe score from the publisher.IndexThis result as a score out of 100.RunThe settings of the run.DateDate of the result.Published byThe source of the result.
MATH level 5Math97.8%46.8high effort—Epoch AI
MATH 500Math94.2%47.1—9 Jan 2026Vals AI
MGSMMultilingual93.4%——9 Jan 2026Vals AI
AIMEMath83.7%52.7—16 Apr 2026Vals AI
LiveCodeBenchCoding82.2%59.1—1 Sept 2026Vals AI
OTIS Mock AIME 2024-2025Math81.7%57.2high effort—Epoch AI
MMLU ProKnowledge80.6%47.4—1 Sept 2026Vals AI
MMMU ProMultimodal79.7%57.0—1 Sept 2026Vals AI
GPQA diamondKnowledge79.6%51.7high effort—Epoch AI
GPQA DiamondKnowledge74.5%47.0—1 Sept 2026Vals AI
τ²-bench RetailAgentic68.3%47.8Sierra30 Apr 2026Sierra Research
ARC-AGI-1 (semi-private)Reasoning58.7%54.9high effort—ARC Prize Foundation
SWE-bench VerifiedCoding54.8%41.5PatchPilot-v1.11 Sept 2026SWE-bench team
τ²-bench AirlineAgentic52.1%36.2Sierra2 Mar 2026Sierra Research
τ²-bench TelecomAgentic50.2%34.8Sierra2 Mar 2026Sierra Research
SWE-bench VerifiedCoding45.0%33.6mini-SWE-agent26 Feb 2026SWE-bench team
FrontierMath-Tiers-1-3-v2-PrivateMath36.1%51.8high effort—Epoch AI
SWE-bench MultimodalMultimodal33.9%—GUIRepair17 Nov 2025SWE-bench team
SWE-bench MultimodalMultimodal33.8%—GUIRepair17 Nov 2025SWE-bench team
Chess PuzzlesReasoning26.0%57.1high effort—Epoch AI
FrontierMath-2025-02-28-PrivateMath24.8%53.8high effort—Epoch AI
SimpleQA VerifiedKnowledge18.8%38.9high effort—Epoch AI
FrontierMath-Tier-4-2025-07-01-PrivateMath6.3%50.2high effort—Epoch AI
ARC-AGI-2 (semi-private)Reasoning6.1%42.8high effort—ARC Prize Foundation
Mystery Game PuzzlesReasoning5.0%39.2high effort—Epoch AI
FrontierMath-Tier-4-v2-PrivateMath4.9%50.3high effort—Epoch AI
IOI v1Coding4.8%42.7—9 Aug 2026Vals AI

20 benchmarks count, from 24 of 27 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 AISierra Research, collected directlyMIT — results are in the licensed repositoryARC Prize Foundation, collected directlyNo licence stated. Their terms ask for written permission before commercial useSWE-bench team, collected directlyNo licence stated. The repository publishes submission records for reproducibility and transparency and asks that SWE-bench be cited

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