# Gemini 3.6 Flash: AI model fact sheet

- **Provider:** google
- **Released:** 2026-07-21
- **Training cutoff:** Mar 2026
- **Context window:** 1,048,576 tokens
- **Max output:** 65,536 tokens
- **API pricing:** $1.50 / 1M input, $7.50 / 1M output
- **OpenRouter ID:** google/gemini-3.6-flash
- **Capabilities:** conversation, reasoning, code-generation, analysis, tool-use, agentic-tool-use, function-calling

Gemini 3.6 Flash is Google's high-efficiency workhorse model, tuned for coding, agentic workflows, and web and app development. Built on Gemini 3.5 Flash, it aims for polished output with fewer unnecessary edits and less hedging, spending roughly 17% fewer output tokens and fewer tool calls to finish the same multi-step task. It accepts text, images, audio, video, and PDFs across a 1M-token context window and returns up to 64K tokens of text, with configurable reasoning effort, structured output, and tool use.

## Benchmarks

| Benchmark | Score |
| --- | --- |
| SWE-bench Pro | 58.7% |
| Terminal-Bench 2.1 | 78.0% |
| DeepSWE v1.1 | 49.0% |
| MLE-Bench | 63.9% |
| OSWorld-Verified | 83.0% |
| CharXiv | 85.2% |
| GDM-MRCR v2 (128k) | 91.8% |

Source: real side-by-side outputs, pricing and specs at https://rival.tips/models/gemini-3.6-flash