Skip to content
aiqbench

OpenAI · model passport

OpenAI: GPT-4.1

gpt-4-1
  • Cloud only
  • Proprietary licence
  • 1,047,576 context
  • Released 2025-04-14(curated)
20/27axes covered
Overall · LMArena
1,415#130 of 396
Best percentile
81.0HHEM Hallucination (Vectara)
Cohort standing
47.0median 48.1

What the data says

Benchmarked on 20 of 27 tracked axes. Every figure below is the model's own published result on that board.

Weakest axes

  • 6.3 pct — SWE-bench Verified (48.5%)
  • 13.5 pct — ARC-AGI-2 (0.4%)

Uneven: 74.7 percentile points between its best and worst axis — 81 pct on HHEM Hallucination (Vectara) versus 6.3 pct on SWE-bench Verified.

  • HHEM Hallucination (Vectara)81 pct94.4%Verify →
  • BFCL v4 (function calling)Dormant board80.5 pct54.0%Verify →
  • MMLU-Pro (Artificial Analysis)69.7 pct80.6%Verify →
  • Instruction following68.2 pct1,403Verify →
  • Overall67.4 pct1,415Verify →
  • Hard prompts67.2 pct1,432Verify →
  • Coding65.5 pct1,456Verify →
  • LiveBench59.8 pct58.4%Verify →
  • LiveCodeBench (Artificial Analysis)53.4 pct45.7%Verify →
  • Vision Arena51.3 pct1,214Verify →
  • GPQA Diamond48.4 pct66.9%Verify →
  • IFBench (Artificial Analysis)47.5 pct43.0%Verify →
  • AIME (math)35.5 pct38.3%Verify →
  • FrontierMath34.8 pct5.5%Verify →
  • SimpleQA Verified26.8 pct31.1%Verify →
  • Factuality Arena20.9 pct1,408Verify →
  • SimpleBench18.4 pct27.0%Verify →
  • Humanity's Last Exam15.2 pct5.4%Verify →
  • ARC-AGI-213.5 pct0.4%Verify →
  • SWE-bench Verified6.3 pct48.5%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
  • 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/gpt-4.1
Context length1,047,576 tokens
Max output32,768 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

Max completion tokensMax tokensResponse formatSeedStructured outputsTemperatureTool choiceTool callingTop-p

Example request

curl https://openrouter.ai/api/v1/chat/completions \
  -H "Authorization: Bearer $OPENROUTER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"openai/gpt-4.1","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.

  • gpt-4.1-2025-04-14 stands at 47 (raw mean 46.6) — measured on 20 of 27 axes.

    standingtemplateVerify →

  • gpt-4.1-2025-04-14 stands at 47 versus a cohort median of 48.1 across 151 models released within 180 days — below 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 →