Gemini 3.6 Flash
2026-07-21 · Publisher : Google · Domain : Artificial Intelligence
Google launches Gemini 3.6 Flash, a workhorse model that is more token-efficient than 3.5 Flash, cheaper on output and better at code, business knowledge and computer use, alongside 3.5 Flash-Lite and 3.5 Flash Cyber.
What changes
- 17% fewer output tokens than 3.5 Flash according to the Artificial Analysis index, up to 65% on DeepSWE.
- Gains in code (DeepSWE 49% versus 37%), machine learning (MLE-Bench) and business knowledge (GDPval-AA v2).
- Computer use built in as a client-side tool in the Gemini API and Gemini Enterprise.
- Same input price as 3.5 Flash but cheaper output.
- Released at the same time as 3.5 Flash-Lite and 3.5 Flash Cyber (a cybersecurity model associated with the CodeMender agent, restricted access).
Google launches Gemini 3.6 Flash, a workhorse model that is more token-efficient than 3.5 Flash, cheaper on output and better at code, business knowledge and computer use, alongside 3.5 Flash-Lite and 3.5 Flash Cyber. ### What changes - 17% fewer output tokens than 3.5 Flash according to the Artificial Analysis index, up to 65% on DeepSWE. - Gains in code (DeepSWE 49% versus 37%), machine learning (MLE-Bench) and business knowledge (GDPval-AA v2). - Computer use built in as a client-side tool in the Gemini API and Gemini Enterprise. - Same input price as 3.5 Flash but cheaper output. - Released at the same time as 3.5 Flash-Lite and 3.5 Flash Cyber (a cybersecurity model associated with the CodeMender agent, restricted access). ### Specifications - Context: up to 1 million tokens; output up to 64,000 tokens; text, image, audio and video inputs. - Knowledge cutoff: March 2026. - Availability: Gemini API, AI Studio, Antigravity, Gemini Enterprise Agent Platform, Gemini app. - API identifier: gemini-3.6-flash. ### Pricing | Item | Price (USD / 1M tokens) | |---|---| | Input (announcement) | 1.50 | | Output (announcement) | 7.50 | The current pricing page shows 0.75 for input and 3.75 for output until December 31, 2026, then 1.50 and 7.50 from January 1, 2027 (introductory cut aligned with 3.7 Flash). ### Benchmarks | Benchmark (scores published by the vendor) | 3.6 Flash | 3.5 Flash | Claude Sonnet 5 | |---|---|---|---| | SWE-Bench Pro | 58.7% | 55.1% | 63.2% | | DeepSWE v1.1 | 49% | 37% | 54% | | Terminal-Bench 2.1 | 78.0% | 76.2% | 80.4% | | MLE-Bench | 63.9% | 49.7% | 66.9% | | OSWorld-Verified | 83.0% | 78.4% | 81.2% | | GDPval-AA v2 (Elo) | 1421 | 1349 | 1607 | ### Key takeaways An update targeted at cost per agentic task; to be considered intermediate, as 3.7 Flash came out three weeks later.
Links
Tags : LLM, Google, Gemini, Flash, Agents