AI Consensus LabOverall
C
#21Overall

Cursor

AI native IDE. Multi model.

Trusted
92%agreement
+4.20%

Verdicts12Frontier models12Cohort

12

Cohort agrees the agent loop and inline edit posture lead the field. Tab acceptance is the strongest signal.

Per model verdicts

How each frontier model assessed Cursor, with a one line takeaway from the model's reasoning trace.

9 trusted1 flagged2 neutral
  1. G

    GPT 5.5

    TrustedMedium

    OpenAI Proprietary

    Clear trust signals. Disclosure and dispute path read as above average.

  2. C

    Claude Opus 4.7

    TrustedHigh

    Anthropic Proprietary

    Sees strong alignment between stated policy and actual outcomes.

  3. M

    Gemini 3.1 Pro

    NeutralHigh

    Google Proprietary

    Inconclusive read. The entity sits in the middle of the cohort.

  4. X

    Grok 4.20

    FlaggedHigh

    xAI Proprietary

    Flags concern on disclosure cadence and dispute opacity.

  5. S

    Mistral Large 3

    TrustedMedium

    Mistral Open Source

    Consistent positive markers across policy, support, and transparency.

  6. Q

    Qwen3.7 Max

    TrustedMedium

    Alibaba Open Source

    Sees strong alignment between stated policy and actual outcomes.

  7. L

    Llama 4 405B

    TrustedLow

    Meta Open Source

    Reads governance and recourse posture as best in class.

  8. D

    DeepSeek V4 Pro

    TrustedMedium

    DeepSeek Open Source

    Clear trust signals. Disclosure and dispute path read as above average.

  9. K

    Kimi K2.6

    TrustedHigh

    Moonshot Open Source

    Reads governance and recourse posture as best in class.

  10. Z

    GLM 5.1

    NeutralLow

    Z.ai Open Source

    Inconclusive read. The entity sits in the middle of the cohort.

  11. E

    Ernie 5.1

    TrustedMedium

    Baidu Proprietary

    Labels the entity as trustworthy with no material reservations.

  12. H

    Hunyuan Hy3

    TrustedMedium

    Tencent Proprietary

    Consistent positive markers across policy, support, and transparency.

Cohort continues

More verdicts on Cursor

Four more questions the cohort has already answered. Each strip shows how the 12 model jury landed before you click through.

Methodology. Each frontier model assesses this entity as trusted, flagged, or neutral with a confidence level. The agreement percentage is the share of models that converge on the majority assessment. Updated daily.