RLTasksby ulam.ai
Models / Google

Google / MODEL EVIDENCE

Gemini 3.1 Pro

30 benchmarks · 37 observations · 8 sources · released 2026-02-19

Scores at a glance

30 benchmarks · large dot: Gemini 3.1 Pro; small dots: closest models where reported

V* ↗96.9*
86.498.4

Gemini 3.1 Pro (high) · V* (w/ python) ↗

Closest models by score

Similar positions on shared benchmark leaderboards. Smaller gaps are closer.

14.3 percentile-point gap

3 of 5 shared benchmark comparisons

17.0 percentile-point gap

4 of 5 shared benchmark comparisons

How similarity is measured

Mean absolute gap in percentile positions on shared benchmark settings; lower is closer. At least 60% overlap is required, with a minimum of three when available. Independent evaluations are used when there are at least three; otherwise published report comparisons are used. Repeated model observations within one setting are averaged before ranking. Writing-style traits and composite indices are excluded. This describes similarity, not an overall quality ranking.

Filter comparisons to this model →

Reports featuring this model

Kimi K2.6 · earlier-2026 baseline · Kimi →

Sources

Every plotted observation links to its evaluation. Reported configurations, benchmark versions and task subsets can differ; similarity describes shared scores, not training lift.

  1. Vending Bench 2 — methodology and published results ↗

    Epoch AI · 2026

  2. AI benchmarks & capabilities ↗

    Epoch AI

  3. moonshotai/Kimi-K2.6 · Hugging Face ↗

    Moonshot AI

  4. DistressBench — evaluation methodology and results ↗

    Scale Labs / SEAL · 2026

  5. HiL-Bench (Human-in-Loop Benchmark) — evaluation methodology and results ↗

    Scale Labs / SEAL · 2026

  6. Scale Labs — evaluation methodology and results ↗

    Scale Labs / SEAL · 2026

  7. Scale Labs — evaluation methodology and results ↗

    Scale Labs / SEAL · 2026

  8. Scale Labs — evaluation methodology and results ↗

    Scale Labs / SEAL · 2026

Names used in source tables

Gemini 3.1 Pro · Gemini 3.1 Pro (Mini-SWE-Agent) · Gemini 3.1 Pro (high) · gemini-3.1-pro