Sponsored byFusecodeEnterprise coding API for Claude Code, Codex, and model workflows.

OpenAI

o4 Mini High API Benchmarks, Pricing & Provider Data

OpenAI o4-mini-high is the same model as [o4-mini](/openai/o4-mini) with reasoning_effort set to high. OpenAI o4-mini is a compact reasoning model in the o-series, optimized for fast, cost-efficient p...

Quality

#62of 96

44.0

LMSpeed score

Category Performance

Scores across benchmark categories on a 0-100 scale.

Coverage: 0 / 8

Avg. score

-

No benchmark category data yet.

No data:AgentsCodingReasoningKnowledgeMath

Specifications

Input and output token limits for this model, plus how it ranks on long-context understanding.

INPUT

200Ktokens

≈ 240 pages of text

OUTPUT

100Ktokens

8K128K1M4M

200K

Features

Technical Details

Input

Detailed scores

Updated: Jul 9, 2026

Overall

1 metric

Overall score44.0#62 / 96

Overall score

LMSpeed rank: #62 / 96Updated: Jul 9, 2026Confidence: 1

6047.0

6244.0

62o4 Mini High44.0

6542.0

6740.0

6939.0

Reasoning & knowledge

4 metrics

Reasoning score46.8#41 / 65Math score29.6#29 / 29

Reasoning score

LMSpeed rank: #41 / 65Updated: Jul 9, 2026Confidence: 1

3750.1

38

Coding & agents

2 metrics

Coding score48.6#58 / 83Agentic score44.8#51 / 81

Coding score

LMSpeed rank: #58 / 83Updated: Jul 9, 2026Confidence: 1

5454.0

55

Language & media

2 metrics

Multilingual score67.2#22 / 59Multimodal Grounded score49.5#53 / 70

Multilingual score

LMSpeed rank: #22 / 59Updated: Jul 9, 2026Confidence: 1

1970.6

19

Related model comparisons

GGPT-5.4 vs o4 Mini High

3 verified comparison points

Indexable

GGPT-5 vs o4 Mini High

3 verified comparison points

Indexable

CClaude Opus 4.6 vs o4 Mini High

3 verified comparison points

Indexable

Multilingual

Multimodal

Instruction following

Output

Knowledge cutoff

2024-06-30

Documentation

Tokenizer

GPT

Architecture

text+image+file->text

Moderated

Yes

Supported parameters

include_reasoningmax_tokensreasoningresponse_formatseedstructured_outputstool_choicetools

Mistral Large 3

49.2

39Llama 3.147.9

40Claude 3.5 Sonnet47.2

41o4 Mini High46.8

42GPT-4o44.4

43Qwen3.6 Plus43.6

44Gemini 2.5 Flash42.2

45Claude 3 Opus41.3

46Llama 4 Scout39.7

Math score

LMSpeed rank: #29 / 29Updated: Jul 9, 2026Confidence: 1

20Gemini 3 Pro63.9

21Claude Opus 4.661.6

22Kimi K2.558.7

23Gemini 3 Flash56.1

24GPT-548.8

2548.6

2644.1

2730.3

2830.0

29o4 Mini High29.6

Knowledge score

LMSpeed rank: #47 / 79Updated: Jul 9, 2026Confidence: 1

43MiMo-V2-Flash58.6

44O155.5

45O3 Mini55.2

46Grok 3 Mini54.5

47o4 Mini High54.2

4853.8

4953.4

5053.0

5150.7

5250.6

Instruction Following score

LMSpeed rank: #39 / 66Updated: Jul 9, 2026Confidence: 1

38GPT-4.1 Nano62.9

39Claude 3.5 Sonnet62.4

39Claude Opus 4.162.4

39Claude Sonnet 462.4

39o4 Mini High62.4

4362.1

4461.6

4561.2

4660.1

4759.9

DeepSeek V3.2

53.6

56O153.1

57Grok 3 Mini49.0

58o4 Mini High48.6

59Llama 3.146.5

60Gemini 2.5 Pro46.1

61Gemini 3 Flash45.5

62Claude 3.5 Sonnet44.7

63Mistral Large 336.7

Agentic score

LMSpeed rank: #51 / 81Updated: Jul 9, 2026Confidence: 1

47Kimi K2.550.3

48GLM-5V Turbo48.8

49Qwen3.6 35B A3B48.2

50Hy3 preview47.7

51o4 Mini High44.8

5244.1

5343.9

5442.1

5540.8

Qwen3.5-27B

70.6

21Gemini 2.5 Pro70.0

22Claude Sonnet 467.2

22o4 Mini High67.2

22Qwen3.5-35B-A3B67.2

25Claude Opus 4.164.4

25O364.4

25O3 Pro64.4

28Qwen362.7

Multimodal Grounded score

LMSpeed rank: #53 / 70Updated: Jul 9, 2026Confidence: 1

49GLM-556.5

50Claude 3 Haiku56.3

51Gemini 2.5 Flash54.8

52GPT-5.4 Nano53.9

53o4 Mini High49.5

5449.3

5545.8

5645.6

5743.6

5840.7

Gemini 3.1 Flash Lite