Skip to content
aiqbench

OpenAI · model passport

OpenAI: GPT-6 Astra

gpt-6-astra
  • Cloud only
  • Proprietary licence
  • 1,050,000 context
  • Released 2026-09-13(first seen on LMArena)
17/27axes covered
Overall · LMArena
1,477#27 of 396
Best percentile
100.0AIME (math)
Cohort standing
85.2median 60.4

What the data says

Benchmarked on 17 of 27 tracked axes. Every figure below is the model's own published result on that board.

Strongest axes

  • 100 pct — AIME (math) (100.0%)
  • 100 pct — ARC-AGI-2 (95.0%)
  • 100 pct — Humanity's Last Exam (54.8%)

Uneven: 41.9 percentile points between its best and worst axis — 100 pct on AIME (math) versus 58.1 pct on HHEM Hallucination (Vectara).

  • AIME (math)100 pct100.0%Verify →
  • ARC-AGI-2100 pct95.0%Verify →
  • Humanity's Last Exam100 pct54.8%Verify →
  • GPQA Diamond99.5 pct95.8%Verify →
  • WebDev Arena99.3 pct1,780Verify →
  • SimpleQA Verified98.8 pct75.6%Verify →
  • Coding98.5 pct1,543Verify →
  • Agent Arena96.4 pct0.141Verify →
  • Search Arena95.2 pct1,246Verify →
  • Instruction following93.9 pct1,473Verify →
  • Hard prompts93.7 pct1,502Verify →
  • Overall93.4 pct1,477Verify →
  • Vision Arena90 pct1,286Verify →
  • Terminal-Bench 4.087 pct57.9%Verify →
  • Factuality Arena82.6 pct1,466Verify →
  • Document Arena68.4 pct1,473Verify →
  • HHEM Hallucination (Vectara)58.1 pct91.3%Verify →
  • FrontierMathNot benchmarked
  • SimpleBenchNot benchmarked
  • LiveBenchNot benchmarked
  • BFCL v4 (function calling)Not benchmarked
  • SWE-bench Verified (live board)Not benchmarked
  • MMLU-Pro (Artificial Analysis)Not benchmarked
  • LiveCodeBench (Artificial Analysis)Not benchmarked
  • IFBench (Artificial Analysis)Not benchmarked
  • SWE-bench VerifiedNot 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-6-astra
Context length1,050,000 tokens
Max output128,000 tokens
Acceptsfile, image, text
Input price$10 per 1M tokens
Cached input price$1 per 1M tokens
Output price$50 per 1M tokens

API parameters this endpoint supports

Include reasoningMax completion tokensMax tokensReasoningReasoning effortResponse formatSeedStructured outputsTool choiceTool callingVerbosity

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-6-astra","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-6-astra-max stands at 85.2 (raw mean 91.5) — measured on 17 of 27 axes.

    standingtemplateVerify →

  • gpt-6-astra-max stands at 85.2 versus a cohort median of 60.4 across 175 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 →