Gemini 3.1 Pro
Gemini 3.1 Pro is an API-only (closed weights) language model from Google, released February 19, 2026. The context window is 1M tokens. On the Epoch Capabilities Index it scores 156.2, rank 39 of 256 rated models. API price: $2 input / $12 output per million tokens.
Check it against your own PC →
Lab
Google
Released
February 19, 2026
Weights
API-only (closed weights)
Parameters
–
Context
1M
Price / M tokens
$2 / $12 in / out
Epoch ECI
156.2 #39
Strengths by category
ModelTape's 0–100 fit on Epoch AI benchmark results.
| Category | Score | Based on |
|---|---|---|
| Reasoning | 32.8 | Chess Puzzles, SimpleBench, Mystery Game Puzzles, Humanity's Last Exam |
| Math | 45.5 | ProofBench, OTIS Mock AIME 2024-2025, FrontierMath-Tiers-1-3-v2-Private, FrontierMath-Tier-4-v2-Private |
| Coding | 43.4 | GSO-Bench, MirrorCode, Terminal Bench, SWE-bench Verified, DeepSWE |
| Knowledge | 83.7 | GPQA diamond, SimpleQA Verified |
Benchmark results
| Benchmark (Epoch AI) | Score |
|---|---|
| GPQA Diamond | 96.5% |
| OTIS Mock AIME | 95.6% |
| Terminal-Bench | 80.2% |
| SWE-bench Verified | 75.6% |
| SimpleBench | 75.5% |
| SimpleQA Verified | 73.5% |
| FrontierMath (Tiers 1–3) | 62.8% |
| Balrog | 57.0% |
| Chess Puzzles | 48.4% |
| Humanity's Last Exam | 46.4% |
| Mystery Game Puzzles | 27.3% |
| FrontierMath-Tier-4-v2-Private | 26.8% |
| ProofBench | 26.0% |
| GSO-Bench | 22.6% |
| EBR-bench | 14.3% |
| DeepSWE | 11.7% |
| MirrorCode | 8.9% |
| Furniture Assembly | 0.0% |
Links
Announcement · Epoch AI benchmarks
Models with a similar score
GPT-5.5 · Claude Opus 4.8 · DeepSeek V4.1 Flash 1017 · Gemini 3.8 Flash · Kimi K3 · Gemini 3.7 Flash · Muse Spark 1.3 · GPT-5.4
More from Google
Gemini 4 Carbon · Gemini 4 Argon · Gemini 3.8 Flash · Gemini 3.7 Flash · Gemini 3.6 Flash · Gemini 3.5 Flash-Lite · Gemini 3.5 Flash Cyber · Gemma 4 E4B