Skip to content
Model Index
  • Overview
  • Models
  • Compare
  • Benchmarks
  • Finder
LabsMethodology
  1. Benchmarks
  2. Knowledge & reliability

AA-Omniscience

Independent testArtificial Analysis · updated Oct 8, 2026

What it measures

6,000 factual questions across professional domains, scoring both accuracy and how often the model invents an answer instead of abstaining.

How to read it

Runs from −100 to 100. Zero means wrong answers cancel out right ones; negative means the model hallucinates more than it knows. The best single check on trustworthiness.

Open at Artificial Analysis
Best
46
Claude Opus 5.5
Median
1
DeepSeek V4 Pro

Results

39 of 41 tracked models have a result. This scale has no meaningful zero, so bars run from the lowest to the highest tracked score.

Filter in explorer
  1. 1
    Claude Opus 5.5

    Anthropic · Max effort

    46
  2. 2
    Claude Fable 5.1

    Anthropic · Max effort

    43
  3. 2
    GPT-6 Astra

    OpenAI · Max effort

    43
  4. 4
    Gemini 4 ArgonLimited

    Google DeepMind · High effort

    42
  5. 4
    GPT-6.1 Sol

    OpenAI · Max effort

    42
  6. 6
    Claude Sonnet 5.5

    Anthropic · Max effort

    32
  7. 6
    Gemini 3.1 Pro

    Google DeepMind

    32
  8. 6
    Grok 4.7

    SpaceXAI · xHigh effort

    32
  9. 9
    Gemini 3.8 Flash

    Google DeepMind · High effort

    30
  10. 10
    Muse Spark 1.3

    Meta · Max effort

    25
  11. 11
    Kimi K3Open

    Moonshot AI · Max effort

    20
  12. 12
    Step 5 Preview

    StepFun

    16
  13. 13
    GLM-5.3Open

    Z.ai · Max effort

    14
  14. 14
    Qwen3.8-Max

    Alibaba

    12
  15. 15
    Claude Haiku 5.5

    Anthropic · Max effort

    11
  16. 16
    MiMo-V2.6-ProOpen

    Xiaomi

    8
  17. 17
    GLM-5.3-FlashOpen

    Z.ai

    7
  18. 18
    Gemini 3.5 Flash-Lite

    Google DeepMind

    5
  19. 19
    TM
    InklingOpen

    Thinking Machines Lab · xHigh effort

    2
  20. 20
    DeepSeek V4 ProOpen

    DeepSeek · Max effort

    1
  21. 20
    GPT-6 Luna

    OpenAI · Max effort

    1
  22. 20
    MiniMax-M3Open

    MiniMax

    1
  23. 23
    GPT-5.6 Terra

    OpenAI · Max effort

    0
  24. 23
    Nemotron 3 UltraOpen

    NVIDIA · Reasoning effort

    0
  25. 25
    Command A+Open

    Cohere

    -4
  26. 26
    DeepSeek V4.1 FlashOpen

    DeepSeek · Max effort

    -5
  27. 26
    Mistral Large 4

    Mistral AI

    -5
  28. 28
    TM
    Inkling-SmallOpen

    Thinking Machines Lab

    -9
  29. 29
    Kimi K2.7 CodeOpen

    Moonshot AI

    -10
  30. 29
    Qwen3.8-27BOpen

    Alibaba · xHigh effort

    -10
  31. 29
    Qwen3.8-FlashOpen

    Alibaba

    -10
  32. 32
    MiMo-V2.6-FlashOpen

    Xiaomi

    -13
  33. 33
    Hy3Open

    Tencent

    -18
  34. 34
    Mistral Small 4Open

    Mistral AI · Reasoning effort

    -30
  35. 35
    Muse Glimmer 30BOpen

    Meta · High effort

    -33
  36. 36
    Mistral Medium 3.5Open

    Mistral AI

    -37
  37. 37
    Gemma 4 31BOpen

    Google DeepMind · Reasoning effort

    -48
  38. 38
    gpt-oss-120bOpen

    OpenAI · High effort

    -49
  39. 39
    Nova 2 Lite

    Amazon · High effort

    -55
2 tracked models without a result

Hy4 preview, MAI-Thinking-1.

Missing means the evaluator has not published a run, not that the model would score poorly.

Model Index

An independent snapshot of frontier SI models and how they perform on third-party benchmarks.

Data as of Oct 8, 2026

Explore

  • All models
  • Compare
  • Benchmarks
  • Model finder
  • Labs
  • Methodology & sources

Sources

Benchmark results come from Artificial Analysis, Arena, Vals AI and ARC Prize. Specifications and prices come from each lab's own documentation.

Not affiliated with any lab or evaluator. Names and logos are trademarks of their respective owners and are used only to identify them. Scores change often — check the linked source before making a decision that depends on one.

  • Overview
  • Models
  • Compare
  • Benchmarks
  • Finder
  • Overview
  • Models
  • Compare
  • Benchmarks
  • Finder
  • Overview
  • Models
  • Compare
  • Benchmarks
  • Finder