Gemini 3.6 Flash
Google DeepMind's workhorse Gemini model, released July 2026 alongside 3.5 Flash-Lite. Gemini 3.6 Flash improves on 3.5 Flash across coding, knowledge work, and multimodal tasks while using up to 17% fewer output tokens on the Artificial Analysis Index, making it cheaper to run than its predecessor.
Strengths
- Google's workhorse model for agentic and multimodal tasks
- Uses up to 17% fewer output tokens than 3.5 Flash on the Artificial Analysis Index
- 83.0% on OSWorld-Verified, up from 78.4% for the previous generation
- 63.9% on MLE Bench, up from 49.7%
- Available across the Gemini API, AI Studio, Android Studio, Antigravity, and Gemini Enterprise
Caveats
- Proprietary - Google API only
- Context window is not stated on the launch announcement
- A Flash-tier model - the Gemini Pro line targets the hardest reasoning workloads
- Vendor
- License
- Proprietary
- Release Date
- 2026-07-21
- Modalities
- textimageaudiovideo
| Benchmark | Score | Max |
|---|---|---|
| OSWorld-Verified | 83.0 | 100.0 |
| MLE Bench | 63.9 | 100.0 |
| DeepSWE | 49.0 | 100.0 |
| GDPval-AA v2 | 1421.0 | - |
Capabilities
Vision
Audio
Video
Tool Use
Resources
Pricing
$1.50 per 1M input tokens, $7.50 per 1M output tokens. Cheaper in practice than 3.5 Flash thanks to up to 17% lower output-token usage.