GPT-5.1vsGemini 2.5-Pro
Across 14 shared benchmarks, GPT-5.1 leads overall: GPT-5.1 wins 13, Gemini 2.5-Pro wins 1, with 0 ties and an average score difference of +16.02.
GPT-5.1
OpenAI · 2025-11-12 · Reasoning model
Gemini 2.5-Pro
Google Deep Mind · 2025-06-05 · Reasoning model
GPT-5.113 wins(93%)(7%)1 winGemini 2.5-Pro
Benchmark scores
Grouped by capability, sorted by largest gap within each. 14 shared benchmarks.
General Knowledge
GPT-5.1 4/4| Benchmark | GPT-5.1 | Gemini 2.5-Pro | Diff |
|---|---|---|---|
| ARC-AGI | 72.8025 / 65high | 3747 / 65thinking | +35.80 |
| HLE | 42.7038 / 149Thinking High (With Tools + Internet) | 21.6089 / 149thinking | +21.10 |
| ARC-AGI-2 | 17.6032 / 58high | 4.9043 / 58thinking | +12.70 |
| GPQA Diamond | 88.1025 / 175Thinking High (No Tools) | 86.4038 / 175thinking | +1.70 |
Math and Reasoning
GPT-5.1 3/4| Benchmark | GPT-5.1 | Gemini 2.5-Pro | Diff |
|---|---|---|---|
| FrontierMath | 26.7013 / 60Thinking High (With Tools) | 1123 / 60 | +15.70 |
| FrontierMath - Tier 4 | 12.5029 / 80Thinking High (With Tools) | 2.1056 / 80Normal (No Tools) | +10.40 |
| Simple Bench | 53.2010 / 27high | 62.402 / 27thinking | -9.20 |
| AIME2025 | 9428 / 106Thinking High (No Tools) | 8843 / 106thinking | +6 |
Agent Level Benchmark
GPT-5.1 2/2| Benchmark | GPT-5.1 | Gemini 2.5-Pro | Diff |
|---|---|---|---|
| τ²-Bench - Telecom | 95.6014 / 35Thinking High (With Tools) | 5432 / 35thinking + 使用工具 | +41.60 |
| Terminal Bench Hard | 432 / 13Thinking High (With Tools) | 2512 / 13thinking + 使用工具 | +18 |
AI Agent - Information Search
GPT-5.1 1/1| Benchmark | GPT-5.1 | Gemini 2.5-Pro | Diff |
|---|---|---|---|
| BrowseComp | 50.8034 / 43Thinking High (No Tools) | 7.8042 / 43thinking + 使用工具 | +43 |
AI Agent - Tool Usage
GPT-5.1 1/1| Benchmark | GPT-5.1 | Gemini 2.5-Pro | Diff |
|---|---|---|---|
| Terminal Bench 2.0 | 47.6034 / 43Thinking High (With Tools) | 32.6043 / 43thinking + 使用工具 | +15 |
Coding and Software Engineer
GPT-5.1 1/1| Benchmark | GPT-5.1 | Gemini 2.5-Pro | Diff |
|---|---|---|---|
| SWE-bench Verified | 76.3025 / 103high | 67.2063 / 103thinking | +9.10 |
Multimodal Understanding
GPT-5.1 1/1| Benchmark | GPT-5.1 | Gemini 2.5-Pro | Diff |
|---|---|---|---|
| MMMU | 85.402 / 28Thinking High (No Tools) | 829 / 28thinking | +3.40 |
Specs
| Field | GPT-5.1 | Gemini 2.5-Pro |
|---|---|---|
| Publisher | OpenAI | Google Deep Mind |
| Release date | 2025-11-12 | 2025-06-05 |
| Model type | Reasoning model | Reasoning model |
| Architecture | Dense | Dense |
| Parameters | 0.0 | 0.0 |
| Context length | 400K | 1000K |
| Max output | 131072 | 65536 |
API pricing
Prices use DataLearner records when available; missing fields are not inferred.
| Item | GPT-5.1 | Gemini 2.5-Pro |
|---|---|---|
| Text input | 1.25 美元/100万 tokens | 1.25 美元/100 万tokens |
| Text output | 10 美元/100万 tokens | 10 美元/100 万tokens |
| Cache read | 0.125 美元/100万 tokens | 0.125 美元/100 万tokens |
Summary
- GPT-5.1leads in:General Knowledge (4/4), Math and Reasoning (3/4), Agent Level Benchmark (2/2), AI Agent - Information Search (1/1), AI Agent - Tool Usage (1/1), Coding and Software Engineer (1/1), Multimodal Understanding (1/1)
On average across the 14 shared benchmarks, GPT-5.1 scores 16.02 higher.
Largest single-benchmark gap: BrowseComp — GPT-5.1 50.80 vs Gemini 2.5-Pro 7.80 (+43).
Page generated from structured model, pricing and benchmark records. No real-time LLM is used to write the prose.