01
80.6%
Gemini 3.1 Pro
United StatesVerified evidenceLiveCodeBench
91.7%
Aider Polyglot
69.6%
Leaderboard
Top 6 AI models by raw coding evidence.
Coding ranks the published roster by SWE-bench Verified when coverage exists. LiveCodeBench is shown beside it as the mandatory companion signal, and Aider Polyglot appears when verified.
April 7, 2026Review target April 14, 2026
Buyer-facing table
Ranked by SWE-bench Verified
| Rank | Model | SWE-bench VerifiedRank source | LiveCodeBench | Aider Polyglot |
|---|---|---|---|---|
| 01 | Gemini 3.1 Pro United StatesVerified evidence |
United StatesVerified evidenceEditorial investigation
Raw coding view, stripped back
The coding leaderboard keeps the software engineering benchmark front and center, then exposes companion evidence next to it so buyers can spot harness-driven divergence quickly.
Open top coding model"Choose this when you need the highest reasoning ceiling available and can feed it text, images, audio, or video in the same request."
Gemini 3.1 Pro
