Vibeleaderboard
← Benchmark guide

Model results

FrontierSWE

Choose a shared evaluation protocol to compare model scores. Versions, sources and configurations stay separate.

Atlas snapshot captured 10/19/2026, 10:20:28 AM UTC. This is an import timestamp, not the evaluation date. Coverage may be incomplete.

Official source
← All evaluation groups

FrontierSWE

Source
FrontierSWE via Epoch AI
Version
v2
Measurement date
Not reported
Metric
% · Higher is better
Evaluation scope
Harness proximus · Aggregation Mean@5 · effort max · result date not supplied

Positions are withheld because at least one result lacks explicit independent provenance or a documented evaluation scope. Missing metadata is shown as unreported; matching imported metadata is not independent verification of the protocol.

  1. GPT-6 Astra

    openai/gpt-6-astra-20260903

    65.5%

    View modelEvidence aggregated

  2. Claude Opus 5.5

    anthropic/claude-opus-5-5

    62.3%

    View modelEvidence aggregated

  3. Claude Sonnet 5.5

    anthropic/claude-sonnet-5-5

    61.9%

    View modelEvidence aggregated

  4. Claude Fable 5.1

    anthropic/claude-fable-5-1

    56.3%

    View modelEvidence aggregated

  5. Claude Opus 5

    anthropic/claude-opus-5

    52%

    View modelEvidence aggregated

  6. Claude Fable 5

    anthropic/claude-fable-5

    47%

    View modelEvidence aggregated

  7. Claude Haiku 5.5

    anthropic/claude-haiku-5-5

    43.8%

    View modelEvidence aggregated

  8. GPT-5.6 Sol

    openai/gpt-5.6-sol-20260709

    32.2%

    View modelEvidence aggregated

  9. GLM 5.3

    z-ai/glm-5.3-20260816

    30.2%

    View modelEvidence aggregated

  10. Kimi K3

    moonshotai/kimi-k3-20260715

    25.9%

    View modelEvidence aggregated

  11. Muse Spark 1.3

    meta/muse-spark-1.3-20260902

    25.8%

    View modelEvidence aggregated

  12. GLM 5.3 Flash

    z-ai/glm-5.3-flash-20260826

    18.1%

    View modelEvidence aggregated
FrontierSWE model results | VibeLeaderboard