Across 10 shared benchmarks, Gemini 3.0 Flash leads overall: Gemini 3.0 Flash wins 9, Claude Sonnet 4 wins 1, with 0 ties and an average score difference of +20.48.
Google Deep Mind · 2025-12-17 · AI model
Anthropic · 2025-05-23 · Reasoning model
Grouped by capability, sorted by largest gap within each. 10 shared benchmarks.
| Benchmark | Gemini 3.0 Flash | Claude Sonnet 4 | Diff |
|---|---|---|---|
| HLE | 43.5033 / 150thinking + 使用工具 | 5.52141 / 150 | +37.98 |
| ARC-AGI-2 | 33.6027 / 59thinking | 1.3052 / 59 | +32.30 |
| GPQA Diamond | 90.4015 / 175thinking | 68119 / 175 | +22.40 |
| Benchmark | Gemini 3.0 Flash | Claude Sonnet 4 | Diff |
|---|---|---|---|
| Claw Bench | 85.7015 / 29Thinking (With Tools) | 77.8023 / 29Thinking (With Tools) | +7.90 |
| Pinch Bench | 85.2016 / 37Thinking (With Tools) | 80.5022 / 37Thinking (With Tools) | +4.70 |
| Benchmark | Gemini 3.0 Flash | Claude Sonnet 4 | Diff |
|---|---|---|---|
| SWE-bench Verified | 68.7057 / 103thinking | 80.209 / 103parallel_thinking + 使用工具 | -11.50 |
| SWE-Bench Pro - Public | 49.6027 / 38Thinking High (With Tools) | 42.7032 / 38thinking | +6.90 |
| Benchmark | Gemini 3.0 Flash | Claude Sonnet 4 | Diff |
|---|---|---|---|
| AIME2025 | 99.708 / 106thinking + 使用工具 | 3895 / 106 | +61.70 |
| FrontierMath - Tier 4 | 4.2040 / 80Normal (No Tools) | 072 / 80Normal (No Tools) | +4.20 |
| Benchmark | Gemini 3.0 Flash | Claude Sonnet 4 | Diff |
|---|---|---|---|
| τ²-Bench | 90.203 / 40thinking + 使用工具 | 5233 / 40normal + 使用工具 | +38.20 |
| Field | Gemini 3.0 Flash | Claude Sonnet 4 |
|---|---|---|
| Publisher | Google Deep Mind | Anthropic |
| Release date | 2025-12-17 | 2025-05-23 |
| Model type | AI model | Reasoning model |
| Architecture | Dense | Dense |
| Parameters | 0.0 | 0.0 |
| Context length | 2000K | 200K |
| Max output | 65536 | 64000 |
Prices use DataLearner records when available; missing fields are not inferred.
| Item | Gemini 3.0 Flash | Claude Sonnet 4 |
|---|---|---|
| Text input | 0.5 美元/100万 tokens | 3 美元/ 100万tokens |
| Text output | 3 美元/100万 tokens | 15 美元/100万tokens |
| Cache read | 0.05 美元/100万 tokens | 3.75 美元/ 100万tokens |
| Cache write | Not public | 0.3 美元/ 100万tokens |
On average across the 10 shared benchmarks, Gemini 3.0 Flash scores 20.48 higher.
Largest single-benchmark gap: AIME2025 — Gemini 3.0 Flash 99.70 vs Claude Sonnet 4 38 (+61.70).
Page generated from structured model, pricing and benchmark records. No real-time LLM is used to write the prose.
