Claude Fable 5$22.000/MClaude Opus 4.8$11.000/MClaude Opus 4.7$11.000/MClaude Opus 4.6$11.000/MClaude Opus 4.5$33.000/MClaude Sonnet 3.7$6.600/MClaude Opus 3$33.000/MClaude 2.1$12.800/MClaude 2$12.800/MGPT-5.5$12.500/MGPT-5.2$5.425/MGPT-5.2-Codex$5.425/MGPT-5$3.875/MGPT-4.5$97.500/MGPT-4 Turbo Preview$16.000/MGPT-4$39.000/MGPT-4-32k$78.000/Mo3$19.000/Mo3-mini$2.090/Mo4-mini$2.090/Mo1$28.500/Mo1-mini$5.700/Mo1-preview$28.500/MGemini 3.5 Pro$5.000/MGemini 3.1 Pro$5.000/MGemini 3 Pro$5.000/MGemini 2.5 Pro$3.875/MGemini 1.5 Pro$2.375/MGemini 1.0 Ultra$12.000/MGemini 1.0 Pro$0.800/MClaude Fable 5$22.000/MClaude Opus 4.8$11.000/MClaude Opus 4.7$11.000/MClaude Opus 4.6$11.000/MClaude Opus 4.5$33.000/MClaude Sonnet 3.7$6.600/MClaude Opus 3$33.000/MClaude 2.1$12.800/MClaude 2$12.800/MGPT-5.5$12.500/MGPT-5.2$5.425/MGPT-5.2-Codex$5.425/MGPT-5$3.875/MGPT-4.5$97.500/MGPT-4 Turbo Preview$16.000/MGPT-4$39.000/MGPT-4-32k$78.000/Mo3$19.000/Mo3-mini$2.090/Mo4-mini$2.090/Mo1$28.500/Mo1-mini$5.700/Mo1-preview$28.500/MGemini 3.5 Pro$5.000/MGemini 3.1 Pro$5.000/MGemini 3 Pro$5.000/MGemini 2.5 Pro$3.875/MGemini 1.5 Pro$2.375/MGemini 1.0 Ultra$12.000/MGemini 1.0 Pro$0.800/M
BETA
Subscribe
← Back to Index·All OpenAI models
OpenAI
OpenAIReasoningLIVE INDEX

o3 Mini High

text+file->text

Cheaper reasoning model in the o3 generation. Strong on STEM, lower latency than full o3.

o3 Mini High is a reasoning AI model from OpenAI. It costs $1.100 per million input tokens and $4.400 per million output tokens (blended $2.090/M), with a 200K-token context window.

Profile inherited from upstream o3-mini ↗ — this is a hosted variant of the same open-weights model.

Released Jan 2025Modalities textOfficial model page ↗Provider pricing ↗API docs ↗Pricing source ↗Compare with another model →Estimate monthly cost →
INPUT
$1.100/M
per million input tokens
OUTPUT
$4.400/M
per million output tokens
BLENDED 70/30
$2.090/M
default reference rate · how it's calculated →
CONTEXT
200K
200,000 tokens
What it's good at
  • Cheap reasoning
  • STEM-tuned
  • Lower latency than o3
Typical use cases
  • Reasoning at scale
  • STEM tutoring
  • Cost-sensitive code review
Benchmarks
vs. best public score
Scores inherited from o3-mini — this is a hosted variant of the same open-weights model, so the underlying benchmark scores are identical.
MMLU86%
Multitask academic knowledge across 57 subjects.
GPQA Diamond79%
Graduate-level science questions, "Google-proof".
MATH95%
High-school competition math problems.
HumanEval92%
Python function synthesis from docstrings.
SWE-bench Verified49%
Real GitHub issues solved end-to-end.
LMArena Elo1305 Elo
Crowd-sourced head-to-head preference Elo rating.
Hand-curated from each provider's published reports and public leaderboards. Methodology varies across sources — treat as directional rather than authoritative.
More from OpenAI
See all 25 →
GPT-5.5
Frontier · 1M ctx
in $5.000/Mout $30.000/M
GPT-5.2
Frontier · 400K ctx
in $1.750/Mout $14.000/M
GPT-5.2-Codex
Coding · 400K ctx
in $1.750/Mout $14.000/M
GPT-5
Frontier · 400K ctx
in $1.250/Mout $10.000/M
GPT-4.1
Frontier · 1M ctx
in $2.000/Mout $8.000/M
GPT-4.1 mini
Efficient · 1M ctx
in $0.400/Mout $1.600/M
Frequently asked questions

How much does o3 Mini High cost?

o3 Mini High costs $1.100 per million input tokens and $4.400 per million output tokens, for a blended reference rate of $2.090 per million tokens.

What is o3 Mini High's context window?

o3 Mini High supports up to 200K tokens of context (200,000 tokens).

What is o3 Mini High best for?

o3 Mini High is well suited to Cheap reasoning, STEM-tuned and Lower latency than o3.

Who makes o3 Mini High?

o3 Mini High is developed and served by OpenAI. It was released in Jan 2025.