Skip to content
aiqbench

OpenAI · model passport

OpenAI: o3

o3
  • Cloud only
  • Proprietary licence
  • 200,000 context
  • Released 2025-04-16(curated)
19/27axes covered
Overall · LMArena
1,432#100 of 396
Best percentile
100.0LiveBench
Cohort standing
63.0median 48.1

What the data says

Benchmarked on 19 of 27 tracked axes. Every figure below is the model's own published result on that board.

Strongest axes

  • 100 pct — LiveBench (80.7%)
  • 92.7 pct — BFCL v4 (function calling) (63.0%)
  • 92.5 pct — MMLU-Pro (Artificial Analysis) (85.3%)

Uneven: 81.2 percentile points between its best and worst axis — 100 pct on LiveBench versus 18.8 pct on SWE-bench Verified.

  • LiveBench100 pct80.7%Verify →
  • BFCL v4 (function calling)Dormant board92.7 pct63.0%Verify →
  • MMLU-Pro (Artificial Analysis)92.5 pct85.3%Verify →
  • LiveCodeBench (Artificial Analysis)91.2 pct80.8%Verify →
  • IFBench (Artificial Analysis)84.6 pct71.4%Verify →
  • Overall75 pct1,432Verify →
  • Hard prompts70.5 pct1,441Verify →
  • Coding68.3 pct1,460Verify →
  • Instruction following67.7 pct1,403Verify →
  • SimpleQA Verified67.1 pct49.4%Verify →
  • AIME (math)62.5 pct84.4%Verify →
  • GPQA Diamond62 pct81.8%Verify →
  • Humanity's Last Exam58.7 pct20.3%Verify →
  • SimpleBench55.2 pct53.1%Verify →
  • FrontierMath55.1 pct18.7%Verify →
  • Vision Arena53.3 pct1,216Verify →
  • ARC-AGI-241.6 pct6.5%Verify →
  • Factuality Arena19.2 pct1,404Verify →
  • SWE-bench Verified18.8 pct62.3%Verify →
  • WebDev ArenaNot benchmarked
  • Search ArenaNot benchmarked
  • Document ArenaNot benchmarked
  • Agent ArenaNot benchmarked
  • SWE-bench Verified (live board)Not benchmarked
  • Terminal-Bench 4.0Not benchmarked
  • HHEM Hallucination (Vectara)Not benchmarked
  • Terminal-BenchNot benchmarked

Percentiles are the model's standing within each board's own population. Axes marked "not benchmarked" have no published result for this model — that is an absence of data, not a low score.

How to run it

Use it hosted

API model IDopenai/o3
Context length200,000 tokens
Max output100,000 tokens
Acceptsimage, text, file
Input price$2 per 1M tokens
Cached input price$0.50 per 1M tokens
Output price$8 per 1M tokens

API parameters this endpoint supports

Include reasoningMax tokensReasoningResponse formatSeedStructured outputsTool choiceTool calling

Example request

curl https://openrouter.ai/api/v1/chat/completions \
  -H "Authorization: Bearer $OPENROUTER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"openai/o3","messages":[{"role":"user","content":"Hello"}]}'

View on OpenRouter →

Identifiers, limits and prices are synced from the upstream source shown above. Any field we haven't synced is left out entirely rather than guessed.

  • o3-2025-04-16 stands at 63 (raw mean 65) — measured on 19 of 27 axes.

    standingtemplateVerify →

  • o3-2025-04-16 stands at 63 versus a cohort median of 48.1 across 151 models released within 180 days — above the peer median.

    cohorttemplateVerify →

Every claim is a computed fact from the current board data, phrased by a low-temperature model call and rejected by a validation gate if any number is not in the source row. How the leaderboards work →