Skip to content
Yardstick
  1. Labs
  2. OpenAI

OpenAI

San Francisco, USA

The maker of ChatGPT. Its current GPT-6 generation is sold in named tiers, from the flagship Astra through Sol to the small, fast Luna.

openai.com (opens in a new tab)

Founded
2015
Models listed
3The GPT models and ChatGPT
Latest release
Sep 2026GPT-6.1 Sol
Benchmarks led
7Vending-Bench, ARC-AGI-2, ARC-AGI-3, FrontierMath, MathArena, HLE, GPQA

Line-up

Ranked by Artificial Analysis Intelligence Index. Select a model for its full benchmark profile.

  1. Rank 1
    GPT-6 Astra52.7

    OpenAI · $10 / $50New

  2. Rank 2
    GPT-6.1 Sol51.8

    OpenAI · $2 / $10New

  3. Rank 3
    GPT-6 Luna38.1

    OpenAI · $0.10 / $0.50New

Where it leads

Benchmarks on which one of this lab's models holds the top score in the catalogue.

  • Vending-Bench 2

    $15,515

    GPT-6 Astra

  • ARC-AGI-2

    95%

    GPT-6 Astra

  • ARC-AGI-3

    62.7%

    GPT-6 Astra

  • FrontierMath (Tiers 1–3)

    93.7%

    GPT-6 Astra

  • MathArena

    90.1%

    GPT-6.1 Sol

  • Humanity's Last Exam (Diamond)

    60.6%

    GPT-6 Astra

  • GPQA Diamond

    95.8%

    GPT-6 Astra

Other labs

  • Alibaba
  • Anthropic
  • DeepSeek
  • Google
  • Meta
  • MiniMax
  • Mistral AI
  • Moonshot AI
  • SpaceXAI
  • Xiaomi
  • Z.ai
Yardstick

A plain reading of how today’s SI models compare. Scores are taken from each benchmark’s own leaderboard wherever one exists; anything a lab reported about its own model is marked.

Data as of Oct 8, 2026.

Explore

  • Models
  • Compare
  • Benchmarks
  • Value
  • Labs
  • Methodology & sources

Sources

  • Artificial Analysis (opens in a new tab)
  • Arena (opens in a new tab)
  • Epoch AI Benchmarking Hub (opens in a new tab)
  • ARC Prize (opens in a new tab)
  • Terminal-Bench (opens in a new tab)
  • Scale AI Labs (opens in a new tab)

Yardstick is an independent reference and is not affiliated with, sponsored by or endorsed by any lab listed. Product names and logos are trademarks of their respective owners and are used only to identify them. Benchmarks are imperfect proxies: check a model on your own task before you commit to it.

  • Models
  • Compare
  • Benchmarks
  • Value
  • Labs
  • Models
  • Compare
  • Benchmarks
  • Value
  • Labs
  • Models
  • Compare
  • Benchmarks
  • Value
  • Labs