Skip to main content

Model profile

o3-mini

OpenAIEstablishedReleased Jan 31, 2025
Data verified
Overall Score
47.41Public #136 of 200Verified #70 of 99
Arena Elo
1348
Eligible category ranks
2of 8
Price (1M tokens)
$1.1 in / $4.4 out
API pricing
Speed
160tok/s
Context
200K

Evidence coverage

10 of 323 tracked benchmarks are published. 5 are verified and 5 provisional. 5 of 8 categories are measured.

Updated July 23, 2026Methodology
Published / tracked
10 / 323
Verified
5
Provisional
5
Categories with evidence
5 / 8

Evidence by category

  • Agentic1 benchmark
    Reported
  • Coding2 benchmarks
    Mixed evidence
  • Reasoning0 benchmarks
    Not measured
  • Knowledge5 benchmarks
    Mixed evidence
  • Math1 benchmark
    Verified
  • Multilingual0 benchmarks
    Not measured
  • Multimodal0 benchmarks
    Not measured
  • Inst. Following1 benchmark
    Verified
ProprietaryReasoning
Confidence:
Low
mini

o3-mini ranks #136 out of 200 models on the public leaderboard with an overall score of 47.41/100. It also ranks #70 out of 99 on the verified leaderboard. While not a frontier model, it offers specific advantages depending on the use case.

o3-mini is a proprietary model with a 200K token context window. It uses explicit chain-of-thought reasoning, which typically improves performance on math and complex reasoning tasks at the cost of higher latency and token usage.

o3-mini sits inside the o3 family alongside o3, o3-pro. This profile currently has 10 of 323 tracked benchmarks. BenchLM only exposes non-generated benchmark rows publicly, so missing categories stay blank until a sourced evaluation is available.

Its strongest category is Instruction Following (#9), while its weakest is Coding (#75). This performance profile makes it a well-rounded choice across a range of tasks.

Peer position

Exact provisional scores and ranks for the closest listed peers. A score can appear before a model clears the evidence threshold for a rank, so equal scores can have different rank states.

Range 47.247.74

  1. Claude 3.5 Sonnet
    Anthropic
    #13247.74
    Claude 3.5 Sonnet is #132 with a score of 47.74.
    Compare
  2. GLM-4.5-Air
    Z.AI
    #13347.7
    GLM-4.5-Air is #133 with a score of 47.7.
    Compare
  3. Qwen3.5 Flash
    Alibaba
    #13447.65
    Qwen3.5 Flash is #134 with a score of 47.65.
    Compare
  4. Command A+
    Cohere
    #13547.51
    Command A+ is #135 with a score of 47.51.
    Compare
  5. o3-miniCurrent model
    OpenAI
    #13647.41
    o3-mini is #136 with a score of 47.41.
  6. Gemma 4 12B
    Google
    #13747.29
    Gemma 4 12B is #137 with a score of 47.29.
    Compare
  7. Qwen3.5 Plus
    Alibaba
    #13847.2
    Qwen3.5 Plus is #138 with a score of 47.2.
    Compare

Category percentile

More

Relative position among models eligible for each sourced category. A higher percentile means a stronger position within that category's ranked cohort; 100 is highest.

  1. Inst. Following73%
    Eligible cohort rank #9 of 31Category score 90.3
  2. Coding39%
    Eligible cohort rank #75 of 122Category score 48.3

Category evidence

Scores and ranks appear only where this model has published benchmark evidence. Categories without displayable source records remain not measured.

Category scores, ranks, weighting, benchmark coverage, and evidence status
CategoryScore
AgenticRank Not rankedWeight 22%1 benchmarkReported66.9
CodingRank #75 of 122Percentile 39thWeight 20%2 benchmarksMixed sources48.3
ReasoningWeight 17%0 benchmarksNot measuredNot measured
KnowledgeRank Not rankedWeight 12%5 benchmarksMixed sources65.9
MathWeight 5%1 benchmarkVerifiedScore pending
MultilingualWeight 7%0 benchmarksNot measuredNot measured
MultimodalWeight 12%0 benchmarksNot measuredNot measured
Inst. FollowingRank #9 of 31Percentile 73rdWeight 5%1 benchmarkVerified90.3

Chatbot Arena performance

Scroll horizontally to inspect confidence intervals and vote counts.

Chatbot Arena Elo, confidence interval, and vote count by evaluation view
ViewEloConfidence intervalVotes
Text Overall1348±3.557,313
Coding1416±6.49,455
Math1382±8.54,719
Instruction Following1344±5.116,955
Creative Writing1301±7.08,182
Multi-turn1340±6.69,492
Hard Prompts1370±4.920,100
Hard Prompts (English)1383±6.011,286
Longer Query1358±6.59,224

Benchmark Details

Rows below have a displayable published verification record. Each source link and provenance note remains in the page HTML while its category is closed. Source-unverified manual rows and generated rows stay hidden.

Agentic1 benchmark
τ²-bench resultsReported

τ²-Bench Tool-Agent-User Evaluation

28.7%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
Coding2 benchmarks
SWE-bench VerifiedProvider exact

Software Engineering Benchmark Verified

49.3%Weighted 16%
Source: OpenAI: Introducing GPT-4.1 in the APIProvenance: Provider exact
AA-SciCodeReported

Artificial Analysis SciCode

39.9%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
Knowledge5 benchmarks
GPQAProvider exact

Graduate-Level Google-Proof Q&A

77.2%Weighted 7%
Source: OpenAI: Introducing GPT-4.1 in the APIProvenance: Provider exact
MMLUProvider exact

Massive Multitask Language Understanding

86.9%Display only
Source: OpenAI: Introducing GPT-4.1 in the APIProvenance: Provider exact
Artificial Analysis Intelligence IndexReported
19.0%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
AA-GPQA DiamondReported

Artificial Analysis GPQA Diamond

74.8%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
AA-HLEReported

Artificial Analysis Humanity's Last Exam

8.7%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
Math1 benchmark
AIME 2024Provider exact

American Invitational Mathematics Examination 2024

87.3%Display only
Source: OpenAI: Introducing GPT-4.1 in the APIProvenance: Provider exact
Inst. Following1 benchmark
IFEvalProvider exact

Instruction-Following Eval

93.9%Weighted 35%
Source: OpenAI: Introducing GPT-4.1 in the APIProvenance: Provider exact

o3 Family

Mini

Canonical Entry

o3

Frequently Asked Questions

How does o3-mini perform overall in AI benchmarks?

o3-mini has 10 published benchmark scores on BenchLM, but it does not yet have enough non-generated coverage to receive a global overall rank.

Is o3-mini good for knowledge and understanding?

o3-mini has visible benchmark coverage in knowledge and understanding, but BenchLM does not currently assign it a global category rank there.

Is o3-mini good for coding and programming?

o3-mini ranks #75 out of 122 models in coding and programming benchmarks with an average score of 48.3. There are stronger options in this category.

Is o3-mini good for mathematics?

o3-mini has visible benchmark coverage in mathematics, but BenchLM does not currently assign it a global category rank there.

Is o3-mini good for agentic tool use and computer tasks?

o3-mini has visible benchmark coverage in agentic tool use and computer tasks, but BenchLM does not currently assign it a global category rank there.

Is o3-mini good for instruction following?

o3-mini ranks #9 out of 31 models in instruction following benchmarks with an average score of 90.3. It is among the top performers in this category.

Which sibling models are related to o3-mini?

o3-mini belongs to the o3 family. Related variants on BenchLM include o3, o3-pro.

Does o3-mini have full benchmark coverage on BenchLM?

Not yet. o3-mini currently has 10 published benchmark scores out of the 323 benchmarks BenchLM tracks. BenchLM only exposes non-generated public benchmark rows, so missing categories stay blank until a sourced evaluation is available.

What is the context window size of o3-mini?

o3-mini has a published context window of 200K, which determines how much text it can process in a single interaction.

Last updated: July 23, 2026 · Runtime metrics stay blank until BenchLM has a sourced snapshot.

Choose with this week’s evidence

Join 2,000+ readers for ranking moves, new releases, pricing changes, and the evidence behind them.

Free. One email per week.