> ## Documentation Index
> Fetch the complete documentation index at: https://docs.anyfast.ai/llms.txt
> Use this file to discover all available pages before exploring further.
# Gemini 3.8 Flash
> Use Google Gemini 3.8 Flash through the native Gemini API for long-horizon software engineering, autonomous agents, and complex multimodal workflows.
Gemini 3.8 Flash is Google's generally available Flash model for long-horizon software engineering, autonomous agents, and complex enterprise workflows. AnyFast exposes it through the native Gemini `generateContent` API with the stable model ID `gemini-3.8-flash`.
## Key capabilities
* **1M token context** - Up to 1,048,576 input tokens and 65,536 output tokens
* **Multimodal input** - Text, image, video, audio, and PDF input; text output
* **Long-horizon coding** - Designed for complex, multi-file engineering work and iterative verification
* **Autonomous agents** - Improved multi-step planning, deterministic tool execution, and recovery from failed loops
* **Tools** - Function calling, code execution, Google Search, Google Maps, URL context, and Computer Use (Preview)
* **Structured output** - Generate JSON that follows a supplied response schema
* **Thinking levels** - `low`, `medium`, and `high`; the default is `medium`
* **Caching and processing options** - Supports context caching, Batch API, Flex inference, and Priority inference upstream
## Model specifications
| Property | Value |
| ---------------------- | ---------------------------------- |
| Model ID | `gemini-3.8-flash` |
| Release stage | Generally available (GA) |
| Release date | September 2, 2026 |
| Inputs | Text, image, video, audio, and PDF |
| Output | Text |
| Input token limit | 1,048,576 |
| Output token limit | 65,536 |
| Default thinking level | `medium` |
## Quick example
```bash cURL theme={null}
curl "https://www.anyfast.ai/v1beta/models/gemini-3.8-flash:generateContent?key=YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"contents": [
{
"role": "user",
"parts": [{ "text": "Review this retry design, identify race conditions, and propose a safe implementation plan." }]
}
],
"generationConfig": {
"thinkingConfig": {
"thinkingLevel": "medium"
},
"maxOutputTokens": 2048
}
}'
```
```python Python theme={null}
import requests
response = requests.post(
"https://www.anyfast.ai/v1beta/models/gemini-3.8-flash:generateContent",
params={"key": "YOUR_API_KEY"},
json={
"contents": [{
"role": "user",
"parts": [{
"text": "Review this retry design, identify race conditions, and propose a safe implementation plan."
}],
}],
"generationConfig": {
"thinkingConfig": {"thinkingLevel": "medium"},
"maxOutputTokens": 2048,
},
},
)
response.raise_for_status()
print(response.json()["candidates"][0]["content"]["parts"][0]["text"])
```
## Thinking levels
Choose a thinking level based on latency and task difficulty:
| Level | Recommended use |
| -------- | ----------------------------------------------------------------------------------- |
| `low` | Latency-sensitive chat, drafting, incident response, and quick analysis |
| `medium` | Default; coding and agentic tasks that need balanced latency and reasoning |
| `high` | Difficult coding, mathematics, tool orchestration, and complex multi-step reasoning |
Gemini 3.8 Flash does not support `minimal`. Sending it returns an error. Use `low`, `medium`, or `high`.
Gemini 3.8 Flash may use more reasoning tokens on difficult, long-running tasks because it works in smaller steps, calls tools iteratively, and verifies intermediate results. Use `low` for workflows that do not need this level of verification.
## Verified AnyFast behavior
The following behavior has been verified against the AnyFast international endpoint:
* `low`, `medium`, and `high` thinking levels are accepted; `minimal` returns HTTP 400.
* Text, image, video, audio, and PDF inputs are accepted in one multimodal request.
* Function calling, code execution, Google Search, Google Maps, URL context, Computer Use, structured output, stop sequences, system instructions, and safety settings are accepted.
* File Search is not currently exposed by this endpoint. Sending `tools[].fileSearch` returns HTTP 400.
* Requests ending with a model turn or an empty final user part return HTTP 400.
## Conversation and tool requirements
* Do not prefill the model response. The final conversation turn must be a non-empty `user` turn.
* Preserve every `thoughtSignature` returned by the model when sending conversation history back in a tool workflow.
* A `functionResponse` must include the function `name`. Its `id` and `response` are optional on the current AnyFast route; return the model-provided `id` and the tool result when they are available.
* Keep multimodal assets inside their corresponding content or function-response payload.
## Parameters
| Parameter | Type | Required | Description |
| ----------------------------------------------- | ------- | -------- | --------------------------------------------------------------------------------------------------------------------------------- |
| `key` | string | Yes | API key in the query string |
| `contents` | array | Yes | Conversation turns containing text or multimodal parts |
| `systemInstruction` | object | No | System instruction with a `parts` array |
| `generationConfig.thinkingConfig.thinkingLevel` | string | No | `low`, `medium` (default), or `high` |
| `generationConfig.maxOutputTokens` | integer | No | Maximum output tokens, up to 65,536 |
| `generationConfig.responseMimeType` | string | No | Output MIME type, such as `application/json` |
| `generationConfig.responseSchema` | object | No | JSON Schema used with structured output |
| `generationConfig.stopSequences` | array | No | Strings that stop generation when encountered |
| `tools` | array | No | Function declarations or a supported built-in tool: `googleSearch`, `googleMaps`, `urlContext`, `codeExecution`, or `computerUse` |
| `toolConfig` | object | No | Function-calling or retrieval configuration for a selected tool |
| `safetySettings` | array | No | Per-category content safety thresholds |
View the interactive API reference for Gemini 3.8 Flash.
## Official references
* [Gemini 3.8 Flash model page](https://ai.google.dev/gemini-api/docs/models/gemini-3.8-flash)
* [What's new in Gemini 3.8 Flash](https://ai.google.dev/gemini-api/docs/generate-content/latest-model)
* [Gemini thinking](https://ai.google.dev/gemini-api/docs/thinking)