Opus 4.7vsGemini 3.1 Pro Preview
Across 11 shared benchmarks, Opus 4.7 leads overall: Opus 4.7 wins 6, Gemini 3.1 Pro Preview wins 4, with 1 ties and an average score difference of +2.30.
Opus 4.7
Anthropic · 2026-04-16 · Reasoning model
Across 11 shared benchmarks, Opus 4.7 leads overall: Opus 4.7 wins 6, Gemini 3.1 Pro Preview wins 4, with 1 ties and an average score difference of +2.30.
Anthropic · 2026-04-16 · Reasoning model
Google Deep Mind · 2026-02-20 · Multimodal model
Grouped by capability, sorted by largest gap within each. 11 shared benchmarks.
| Benchmark | Opus 4.7 | Gemini 3.1 Pro Preview | Diff |
|---|---|---|---|
| HLE | 54.706 / 149Extended (with tools) | 51.4012 / 149Thinking High (With Tools) | +3.30 |
| ARC-AGI-2 | 75.809 / 58最高(无工具) | 77.107 / 58Thinking High (No Tools) | -1.30 |
| MMLU | 91.506 / 65Normal (No Tools) | 92.603 / 65Thinking High (No Tools) | -1.10 |
| GPQA Diamond | 94.204 / 175Extended (no tools) | 94.303 / 175Thinking High (No Tools) | -0.10 |
| ARC-AGI-3 | 05 / 6Thinking High (No Tools) | 03 / 6Thinking High (No Tools) | — |
| Benchmark | Opus 4.7 | Gemini 3.1 Pro Preview | Diff |
|---|---|---|---|
| SWE-Bench Pro - Public | 64.302 / 36Extended (with tools) | 54.2017 / 36Thinking High (With Tools) | +10.10 |
| SWE-bench Verified | 87.602 / 103Extended (with tools) | 80.607 / 103Thinking High (With Tools) | +7 |
| Benchmark | Opus 4.7 | Gemini 3.1 Pro Preview | Diff |
|---|---|---|---|
| FrontierMath | 43.806 / 60极高强度思考(无工具) | 36.9011 / 60Thinking High (No Tools) | +6.90 |
| FrontierMath - Tier 4 | 22.9012 / 80极高强度思考(无工具) | 16.7020 / 80Normal (No Tools) | +6.20 |
| Benchmark | Opus 4.7 | Gemini 3.1 Pro Preview | Diff |
|---|---|---|---|
| BrowseComp | 79.3011 / 43Extended (with tools) | 85.903 / 43Thinking High (With Tools + Internet) | -6.60 |
| Benchmark | Opus 4.7 | Gemini 3.1 Pro Preview | Diff |
|---|---|---|---|
| Terminal Bench 2.0 | 69.405 / 43Extended (with tools) | 68.506 / 43Thinking High (With Tools) | +0.90 |
| Field | Opus 4.7 | Gemini 3.1 Pro Preview |
|---|---|---|
| Publisher | Anthropic | Google Deep Mind |
| Release date | 2026-04-16 | 2026-02-20 |
| Model type | Reasoning model | Multimodal model |
| Architecture | Dense | Dense |
| Parameters | 0.0 | 0.0 |
| Context length | 1000K | 1M |
| Max output | 131072 | 32768 |
Prices use DataLearner records when available; missing fields are not inferred.
| Item | Opus 4.7 | Gemini 3.1 Pro Preview |
|---|---|---|
| Text input | $5 / 1M tokens | $2 / 1M tokens |
| Text output | $25 / 1M tokens | $12 / 1M tokens |
| Cache read | $0.5 / 1M tokens | Not public |
| Cache write | $6.25 / 1M tokens | Not public |
On average across the 11 shared benchmarks, Opus 4.7 scores 2.30 higher.
Largest single-benchmark gap: SWE-Bench Pro - Public — Opus 4.7 64.30 vs Gemini 3.1 Pro Preview 54.20 (+10.10).
Page generated from structured model, pricing and benchmark records. No real-time LLM is used to write the prose.