benchrGoogle's release note for July 21, 2026 says Gemini 3.6 Flash and Gemini 3.5 Flash-Lite are now generally available. The company frames 3.6 Flash as a Flash model with better token efficiency, code and agentic planning improvements, and less output verbosity than 3.5 Flash. That last point matters because Flash bills by output tokens, and agentic coding systems often spend more on generated plans, tool calls, and retries than on the prompt that started the run.
The price sheet makes the shape of the update clear. Standard Gemini 3.6 Flash is $1.50 per million input tokens and $7.50 per million output tokens. Gemini 3.5 Flash remains $1.50 input and $9.00 output. Cached input stays $0.15 per million, and cache storage remains $1 per million tokens per hour. Batch and flex cut 3.6 Flash to $0.75 input and $3.75 output. Priority raises it to $2.70 input and $13.50 output.
| Model | Input / 1M | Output / 1M | Cached input / 1M | Best read |
|---|---|---|---|---|
| Gemini 3.6 Flash | $1.50 | $7.50 | $0.15 | New stable Flash default for output-heavy agent loops |
| Gemini 3.5 Flash | $1.50 | $9.00 | $0.15 | Still stable; no shutdown date announced |
| Gemini 3.5 Flash-Lite | $0.30 | $2.50 | $0.03 | Cheaper subagent/router model, not the same capability tier |
What changed for builders
The clean upgrade case is an agent that writes a lot: coding assistants, workflow agents, data-cleaning loops, and systems where each user request fans out into several tool calls. If output is 70% of your bill, a move from $9.00 to $7.50 output cuts the model portion by about 12% before any efficiency gain from shorter answers. If your workload is mostly retrieval, classification, or long input with short labels, the savings are smaller because input did not move.
Capabilities also changed in ways that affect architecture. The model page lists text, image, video, audio, and PDF as supported inputs, text output, caching, code execution, file search, function calling, Maps and Search grounding, structured outputs, thinking, URL context, and preview Computer Use support. That makes 3.6 Flash a broader tool-calling model than the old budget Flash tier. It is not an image generator or Live API voice model.
The migration story is now clearer. Google's deprecations table lists no shutdown date for Gemini 3.6 Flash or 3.5 Flash, but it names 3.6 Flash as the recommended replacement for Gemini 3 Flash Preview, Gemini 2.5 Flash, and several older Flash preview endpoints. If your application still points to a preview Flash model, the work item is straightforward: test 3.6 Flash, measure output-token shrinkage, then pin the stable model ID rather than relying on a moving alias.
What not to overclaim
Google's release note says 3.6 Flash improves token efficiency and code/agentic planning, but benchr is not treating that as a scored benchmark yet. The official model page does not publish a new SWE-bench, GPQA, Terminal-Bench, or ARC-AGI figure for Gemini 3.6 Flash. In the ranking tool, any capability scores are editorial estimates, not official lab results. The hard facts are the release date, stable status, pricing, token limits, and supported features.
FAQ
How much does Gemini 3.6 Flash cost?
Google lists Gemini 3.6 Flash at $1.50 per million input tokens and $7.50 per million output tokens on the standard paid tier, with cached input at $0.15 per million tokens. Batch and flex are $0.75 input and $3.75 output per million tokens.
Is Gemini 3.6 Flash cheaper than Gemini 3.5 Flash?
Only on output. The standard input price stays $1.50 per million tokens, but output drops from $9.00 to $7.50 per million tokens. For output-heavy coding agents, that is a meaningful cut; for retrieval-heavy workloads, the bill may barely move.
Does Gemini 3.6 Flash replace Gemini 3.5 Flash?
Google has not announced a shutdown date for Gemini 3.5 Flash. Gemini 3.6 Flash is the newer stable Flash model, and Google's deprecations page names it as the recommended replacement for several older 2.x and preview Flash models.
Sources
- Google AI for Developers, Gemini API release notes, verified July 22, 2026.
- Google AI for Developers, Gemini API pricing, verified July 22, 2026.
- Google AI for Developers, Gemini 3.6 Flash model page, verified July 22, 2026.
- Google AI for Developers, Gemini deprecations, verified July 22, 2026.
Changelog
July 22, 2026: Published after Google listed Gemini 3.6 Flash and Gemini 3.5 Flash-Lite as generally available in the Gemini API release notes.