Gemini 3.5 Flash

What is Gemini 3.5 Flash?

Gemini 3.5 Flash was the first Gemini 3.5 model, built on Gemini 3 Flash for quick agent loops, sub-agents and iterative coding. It reads text, images, video, audio and PDFs across a 1M-token context and writes only text, up to 65,536 tokens, with four thinking levels that default to medium. It is available only as a hosted model, through the Gemini API, Google AI Studio and Gemini Enterprise Agent Platform.

It is still stable, with no shutdown date, but Google now calls it its legacy Flash model. Gemini 3.6, 3.7 and 3.8 Flash have followed, and they charge $7.50 per million output tokens against its $9.

Is Gemini 3.5 Flash still available?

Yes. It is a stable model with no shutdown date, but Google now calls it its legacy Flash model and points new projects to Gemini 3.8 Flash.

Status

Stable on the Gemini API, with no shutdown date announced.

Successors

Gemini 3.6, 3.7 and 3.8 Flash, all listed above it on Google's pricing page.

Price gap

The newer Flash models charge $7.50 per million output tokens against its $9, with the same $1.50 input.

Gemini app

It launched as the app's default model; Google has since brought Gemini 3.6 Flash to all app users.

Who is Gemini 3.5 Flash best for?

Gemini 3.5 Flash is best for fast agent loops and sub-agents, long documents and media in one prompt, bulk jobs on Batch or Flex, and free prototyping on the Gemini API.

• Sub-agents and agent loops

  Google built it for quick multi-step loops and sub-agent work, with function calling, code execution and structured output built in.

• Long documents and media

  One prompt can hold text, images, video, audio and PDFs up to about 1M tokens.

• Bulk jobs that can wait

  Batch and Flex requests cost half the Standard rate, $0.75 in and $4.50 out per million tokens.

• Prototyping for free

  The free tier covers input and output tokens in Google AI Studio and the API, if you accept that Google may use your prompts to improve its products.

What's included with Gemini 3.5 Flash?

Four thinking levels, built-in tools with Search grounding, computer use in preview, context caching, Batch, Flex and Priority tiers, and reasoning that carries across turns.

Thinking levels

Minimal, low, medium and high trade reasoning depth against speed and cost; medium is the default.

Built-in tools

Function calling, code execution, Google Search and Maps grounding, URL context, file search and structured output.

Computer use (preview)

It can operate a browser or desktop through Google's computer use tool, which is still in preview.

Context caching

Cached input costs $0.15 per million tokens, plus $1 per million tokens per hour of storage.

Batch, Flex and Priority

Batch and Flex halve the price for work that can wait, and Priority costs 1.8 times the Standard rate with its own rate limits.

Thought preservation

Its reasoning carries across the turns of a conversation automatically, which helps multi-step debugging but can raise token use.

What changed from Gemini 3 Flash?

More capable and stable, but three times the price and a lower default thinking level.

Price

$1.50 / $9 per million tokens, against $0.50 / $3 for Gemini 3 Flash Preview.

Status

Generally available and stable, where Gemini 3 Flash stayed a preview.

Thinking

The default level dropped from high to medium, and reasoning now carries across turns.

Scores

Terminal-Bench 2.1 rose from 58.0% to 76.2% and MCP Atlas from 62.0% to 83.6% in Google's tests.

How much does Gemini 3.5 Flash cost?

Gemini 3.5 Flash is free on the Gemini API's rate-limited free tier, and on the paid tier it costs $1.50 per million input tokens and $9.00 per million output tokens, or half that through Batch or Flex.

Prices are for the Gemini Developer API in US dollars, and output prices include thinking tokens. Moving to the paid tier means linking a billing account and prepaying at least $5 of credit; paid prompts are not used to improve Google's products. Grounding with Google Maps also gives 5,000 free prompts a month, then $14 per 1,000. Google AI Studio itself is free to use, and prices on Gemini Enterprise Agent Platform can differ. Full pricing on ai.google.dev. We checked these prices on 28 Sep 2026.

How do readers rate Gemini 3.5 Flash?

No one has rated Gemini 3.5 Flash yet. Be the first.

Gemini 3.5 Flash pros and cons

It is fast, multimodal and free to try, but it costs more than both the Gemini 3 Flash before it and the newer Flash models after it, and it can burn through tokens.

What are the best Gemini 3.5 Flash alternatives?

Five tools on Recaply do Gemini 3.5 Flash's job. DeepSeek V4.1 Flash, Bonsai 27B and Step-Audio-R1 are open source, so you can run them on your own server.

Gemini 3.8 FlashFree

Choose Gemini 3.8 Flash for anything new; it is Google's strongest Flash model, with the same 1M context and free tier, and it costs $1.50 / $7.50 per million tokens against $9 output for 3.5 Flash.

GPT-5.4 mini & nanoPaid

Choose it when you work on OpenAI's API and want half the token price; GPT-5.4 mini costs $0.75 / $4.50 with up to 128K output tokens, but it has a 400K context, no audio or video input and no free tier.

DeepSeek V4.1 FlashOpen Source

Choose it when cost matters most; it charges $0.30 / $1.20 per million tokens at peak hours and half that off-peak, with a 1M context and up to 384K output, though it lists image input but not audio or video.

Bonsai 27BOpen Source

Choose it when you want a 27B model compressed to 1-bit or ternary weights that runs locally on modest hardware.

Step-Audio-R1Open Source

Choose it when audio understanding must run on your own servers instead of sending recordings to Google.

Questions about Gemini 3.5 Flash