Gemini 3.6 Flash is Google's Stable Gemini 3 model for fast, capable generation. Google documents multimodal input, thinking, function calling, and grounding for the underlying model. The current reAPI endpoint exposes a narrower text-only Chat Completions contract: each message `content` must be a string, streaming is supported, and `max_tokens` is capped at 8,192. Current pricing is shown in the table on this page.
Start with the prompt in the composer or choose an example.
Pricing temporarily unavailable
Gemini 3.6 Flash is Google's Stable Gemini 3 model, served here through `https://reapi.ai/api/v1/chat/completions`. Requests use the standard Chat Completions format and are passed to the model as written, so parameters behave the way that format describes. Output is text, and a single response is capped at 8,192 tokens. Use the pricing table on this page for the current input and output rates.
Real-world workflows and production use cases you can build and ship with this model.
Send reports, notes, transcripts, or source excerpts as plain-text string content and ask Gemini 3.6 Flash to summarize, compare, extract decisions, or draft a response. The current endpoint does not upload PDF files or other media, so convert source material to text before sending it. This keeps the request shape simple and auditable.
Read the API docs
Paste code, logs, schemas, or requirements as text and ask for an explanation, refactor, test plan, or structured draft. The endpoint returns text and may display reasoning only when the server returns it. It does not accept `tools`, execute code, or perform grounded web search, so applications should run any external actions themselves.
Set `stream` to `true` to render text chunks as they arrive, or `false` to wait for one JSON response. This fits support assistants, editing flows, classification, and summarization pipelines that work entirely with text. Set `max_tokens` no higher than 8,192 and use the live pricing table on this page to estimate cost.
Gemini 3.6 Flash usage is priced separately for input and output tokens. The pricing table on this page is the source for the current rates; request cost depends on the text tokens actually processed and generated.
The current endpoint is `https://reapi.ai/api/v1/chat/completions`. Send the model, plain-text messages, optional `stream`, and optional `max_tokens`; unsupported multimodal parts, tools, and sampling fields are rejected instead of being silently ignored.
Create or manage the API key in your reAPI account, then authenticate requests to the same reAPI domain. The model-page Playground can also use your signed-in session without exposing a provider credential in the browser.
This compares the underlying Google models using Google's published specifications, not the narrower current reAPI endpoint. Google lists the same limits, inputs, and capability set for both; their published output rates and release recency differ.
Comparison reflects behavior documented on Google's own model and pricing pages at the time of writing. Rates are described as relationships rather than amounts; the live numbers for Gemini 3.6 Flash are in the pricing table above.
Sign in to your current reAPI account and create an API key under Settings. Keep the key server-side and send it as a Bearer token for external API requests.
OpenSet the OpenAI-compatible base URL to `https://reapi.ai/api/v1` and use the key from your reAPI account. The complete endpoint is `https://reapi.ai/api/v1/chat/completions`.
OpenSet `model` to `gemini-3.6-flash`, provide plain-text string `content` in `messages`, and optionally set `stream`. Keep `max_tokens` at or below 8,192; do not send media content parts or `tools` to this endpoint.
Opencurl https://reapi.ai/api/v1/chat/completions \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "gemini-3.6-flash",
"messages": [
{ "role": "user", "content": "Hello" }
],
"stream": true,
"max_tokens": 8192
}'Rejected synchronously with the failing field named. Check enum values, ranges, and URL fields — nothing is charged.
401 means a missing or invalid Bearer key; 402 means the reserve exceeds your balance. Manage keys and credits in the console.
Reference material and outputs pass automated moderation. Rejected tasks fail with a clear error and a full refund.
A task that reaches a failed state is never charged — the reserve refunds automatically. Keep the task id and retry when ready.
Common questions about this model.
Comparisons
Gemini 3.6 Flash vs 3.1 Pro: compare current API prices, the 200K prompt threshold, coding benchmarks, context limits, and the reAPI integration boundary.
Guides
How to use Gemini 3.6 Flash: official benchmarks, where it loses to GPT-5.6 Luna and Sonnet 5, four breaking API changes, and the Flash-Lite and Cyber models.
Explore more models in the same category.
From $0.800 per 1M tokens
OpenAI
From $0.080 per 1M tokens
OpenAI
From $0.160 per 1M tokens
DeepSeek
From $0.400 per 1M tokens
Try it in the playground or grab an API key to integrate now.