Gemini 3.5 FlashvsGPT-5.5
Across 12 shared benchmarks, GPT-5.5 leads overall: Gemini 3.5 Flash wins 2, GPT-5.5 wins 10, with 0 ties and an average score difference of -10.41.
Gemini 3.5 Flash
Google Deep Mind · 2026-06-20 · Multimodal model
GPT-5.5
OpenAI · 2026-04-23 · Reasoning model
Gemini 3.5 Flash2 wins(17%)(83%)10 winsGPT-5.5
Benchmark scores
Grouped by capability, sorted by largest gap within each. 12 shared benchmarks.
AI Agent - Tool Usage
GPT-5.5 2/3| Benchmark | Gemini 3.5 Flash | GPT-5.5 | Diff |
|---|---|---|---|
| MCP-Atlas | 83.604 / 38Thinking High (With Tools) | 75.3021 / 38极高强度思考（工具） | +8.30 |
| Terminal-Bench 2.1 | 76.2022 / 43Thinking High (With Tools) | 83.4012 / 43Thinking High (With Tools) | -7.20 |
| OSWorld-Verified | 78.4010 / 26Thinking High (With Tools) | 78.709 / 26Thinking High (With Tools) | -0.30 |
General Knowledge
GPT-5.5 3/3| Benchmark | Gemini 3.5 Flash | GPT-5.5 | Diff |
|---|---|---|---|
| ARC-AGI-2 | 72.1013 / 62Thinking High (With Tools) | 852 / 62Thinking High (No Tools) | -12.90 |
| HLE | 40.2066 / 181Thinking High (With Tools) | 52.2022 / 181Thinking High (With Tools) | -12 |
| LiveBench | 75.0217 / 115Thinking High (No Tools) | 80.711 / 115Deep Thinking (No Tools) | -5.69 |
Coding and Software Engineer
GPT-5.5 2/2| Benchmark | Gemini 3.5 Flash | GPT-5.5 | Diff |
|---|---|---|---|
| DeepSWE | 3722 / 26Thinking Medium (With Tools) | 677 / 26极高强度思考（工具） | -30 |
| SWE-Bench Pro - Public | 55.1030 / 57Thinking High (With Tools) | 58.6015 / 57Thinking High (With Tools) | -3.50 |
Math and Reasoning
GPT-5.5 2/2| Benchmark | Gemini 3.5 Flash | GPT-5.5 | Diff |
|---|---|---|---|
| FrontierMath Tier 4 v2 | 26.8318 / 34Thinking High (No Tools) | 72.505 / 34极高强度思考（无工具） | -45.67 |
| FrontierMath v2 | 62.8118 / 34Thinking High (No Tools) | 85.266 / 34极高强度思考（无工具） | -22.46 |
Commonsense Reasoning
Gemini 3.5 Flash 1/1| Benchmark | Gemini 3.5 Flash | GPT-5.5 | Diff |
|---|---|---|---|
| SimpleBench | 76.704 / 67Normal (No Tools) | 698 / 67Normal (No Tools) | +7.70 |
General Evaluation
GPT-5.5 1/1| Benchmark | Gemini 3.5 Flash | GPT-5.5 | Diff |
|---|---|---|---|
| GPQA Diamond | 92.8019 / 226Thinking High (No Tools) | 947 / 226极高强度思考（无工具） | -1.20 |
Specs
| Field | Gemini 3.5 Flash | GPT-5.5 |
|---|---|---|
| Publisher | Google Deep Mind | OpenAI |
| Release date | 2026-06-20 | 2026-04-23 |
| Model type | Multimodal model | Reasoning model |
| Architecture | Dense | Dense |
| Parameters | Not available | Not available |
| Context length | 1M | 1000K |
| Max output | 64K | 128K |
API pricing
Prices use DataLearner records when available; missing fields are not inferred.
| Item | Gemini 3.5 Flash | GPT-5.5 |
|---|---|---|
| Text input | $1.5 / 1M tokens | $0.5 / 1M tokens |
| Text output | $9 / 1M tokens | $30 / 1M tokens |
| Cache read | $0.15 / 1M tokens | $0.5 / 1M tokens |
| Cache write | Not public | $6.25 / 1M tokens |
Summary
- Gemini 3.5 Flashleads in:Commonsense Reasoning (1/1)
- GPT-5.5leads in:AI Agent - Tool Usage (2/3), General Knowledge (3/3), Coding and Software Engineer (2/2), Math and Reasoning (2/2), General Evaluation (1/1)
On average across the 12 shared benchmarks, GPT-5.5 scores 10.41 higher.
Largest single-benchmark gap: FrontierMath Tier 4 v2 — Gemini 3.5 Flash 26.83 vs GPT-5.5 72.50 (-45.67).
Page generated from structured model, pricing and benchmark records. No real-time LLM is used to write the prose.