Gemini 3.1 Pro PreviewvsGemini 3.0 Pro (Preview 11-2025)
Across 12 shared benchmarks, Gemini 3.1 Pro Preview leads overall: Gemini 3.1 Pro Preview wins 9, Gemini 3.0 Pro (Preview 11-2025) wins 3, with 0 ties and an average score difference of +8.33.
Google Deep Mind · 2026-02-20 · Multimodal model
Google Deep Mind · 2025-11-18 · Multimodal model
Benchmark scores
Grouped by capability, sorted by largest gap within each. 12 shared benchmarks.
General Knowledge
Gemini 3.1 Pro Preview 3/3| Benchmark | Gemini 3.1 Pro Preview | Gemini 3.0 Pro (Preview 11-2025) | Diff |
|---|---|---|---|
| ARC-AGI-2 | 77.107 / 58Thinking High (No Tools) | 45.1022 / 58parallel_thinking | +32 |
| HLE | 51.4012 / 149Thinking High (With Tools) | 45.8026 / 149high + 使用工具 | +5.60 |
| GPQA Diamond | 94.303 / 175Thinking High (No Tools) | 93.805 / 175parallel_thinking | +0.50 |
Agent Level Benchmark
Gemini 3.1 Pro Preview 2/2| Benchmark | Gemini 3.1 Pro Preview | Gemini 3.0 Pro (Preview 11-2025) | Diff |
|---|---|---|---|
| τ²-Bench | 90.802 / 40Thinking High (With Tools) | 85.408 / 40thinking + 使用工具 | +5.40 |
| τ²-Bench - Telecom | 99.301 / 35Thinking High (With Tools) | 985 / 35high + 使用工具 | +1.30 |
Coding and Software Engineer
Even 2/2| Benchmark | Gemini 3.1 Pro Preview | Gemini 3.0 Pro (Preview 11-2025) | Diff |
|---|---|---|---|
| SWE-bench Verified | 80.607 / 103Thinking High (With Tools) | 76.2027 / 103thinking | +4.40 |
| LiveCodeBench | 91.703 / 118Thinking High (With Tools) | 922 / 118thinking | -0.30 |
Math and Reasoning
Gemini 3.0 Pro (Preview 11-2025) 2/2| Benchmark | Gemini 3.1 Pro Preview | Gemini 3.0 Pro (Preview 11-2025) | Diff |
|---|---|---|---|
| FrontierMath - Tier 4 | 16.7020 / 80Normal (No Tools) | 18.8016 / 80Normal (No Tools) | -2.10 |
| FrontierMath | 36.9011 / 60Thinking High (No Tools) | 3810 / 60thinking | -1.10 |
AI Agent - Information Search
Gemini 3.1 Pro Preview 1/1| Benchmark | Gemini 3.1 Pro Preview | Gemini 3.0 Pro (Preview 11-2025) | Diff |
|---|---|---|---|
| BrowseComp | 85.903 / 43Thinking High (With Tools + Internet) | 59.2029 / 43high + 使用工具 | +26.70 |
AI Agent - Tool Usage
Gemini 3.1 Pro Preview 1/1| Benchmark | Gemini 3.1 Pro Preview | Gemini 3.0 Pro (Preview 11-2025) | Diff |
|---|---|---|---|
| Terminal Bench 2.0 | 68.506 / 43Thinking High (With Tools) | 56.9022 / 43high + 使用工具 | +11.60 |
Claw-style Agent Evaluation
Gemini 3.1 Pro Preview 1/1| Benchmark | Gemini 3.1 Pro Preview | Gemini 3.0 Pro (Preview 11-2025) | Diff |
|---|---|---|---|
| Pinch Bench | 86.7010 / 37Thinking (With Tools) | 70.7031 / 37Thinking (With Tools) | +16 |
Specs
| Field | Gemini 3.1 Pro Preview | Gemini 3.0 Pro (Preview 11-2025) |
|---|---|---|
| Publisher | Google Deep Mind | Google Deep Mind |
| Release date | 2026-02-20 | 2025-11-18 |
| Model type | Multimodal model | Multimodal model |
| Architecture | Dense | Dense |
| Parameters | 0.0 | 0.0 |
| Context length | 1M | 1000K |
| Max output | 32768 | 65536 |
API pricing
Prices use DataLearner records when available; missing fields are not inferred.
| Item | Gemini 3.1 Pro Preview | Gemini 3.0 Pro (Preview 11-2025) |
|---|---|---|
| Text input | $2 / 1M tokens | 2 美元/100万 tokens |
| Text output | $12 / 1M tokens | 12 美元/100万 tokens |
Summary
- Gemini 3.1 Pro Previewleads in:General Knowledge (3/3), Agent Level Benchmark (2/2), AI Agent - Information Search (1/1), AI Agent - Tool Usage (1/1), Claw-style Agent Evaluation (1/1)
- Gemini 3.0 Pro (Preview 11-2025)leads in:Math and Reasoning (2/2)
- Tied in:Coding and Software Engineer
On average across the 12 shared benchmarks, Gemini 3.1 Pro Preview scores 8.33 higher.
Largest single-benchmark gap: ARC-AGI-2 — Gemini 3.1 Pro Preview 77.10 vs Gemini 3.0 Pro (Preview 11-2025) 45.10 (+32).
Page generated from structured model, pricing and benchmark records. No real-time LLM is used to write the prose.