> ## Documentation Index
> Fetch the complete documentation index at: https://docs.bytespike.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# gemini-3-5-flash

> Google's Gemini 3.5 Flash — fast multimodal with a 1M-token context window. The step up from 3-flash for harder reasoning while staying well below Pro pricing.

`gemini-3-5-flash` is Google's mid-tier Gemini — fast and multimodal, with the **1M-token context window** that distinguishes the Gemini family. It offers solid reasoning headroom at a fraction of flagship pricing.

**Pricing:** $1.50 / 1M input, $9.00 / 1M output — see the [rate card](/pricing).

## Protocols

| Protocol                        | Path                                                                           |
| ------------------------------- | ------------------------------------------------------------------------------ |
| Gemini Native                   | `POST https://llm.bytespike.ai/v1beta/models/gemini-3-5-flash:generateContent` |
| OpenAI Chat Completions (shim)  | `POST https://llm.bytespike.ai/v1/chat/completions`                            |
| Anthropic Messages (translated) | `POST https://llm.bytespike.ai/v1/messages`                                    |

For the OpenAI shim, the gateway translates the request body to Gemini's `generateContent` shape behind the scenes. From the client side, you write standard `openai`-SDK code.

## Quickstart

```bash theme={null}
curl https://llm.bytespike.ai/v1/chat/completions \
  -H "Authorization: Bearer $BYTESPIKE_API_KEY" \
  -H "content-type: application/json" \
  -d '{
    "model": "gemini-3-5-flash",
    "messages": [{ "role": "user", "content": "Hello, ByteSpike." }]
  }'
```

## Capabilities

| Capability               | Supported   |
| ------------------------ | ----------- |
| Chat Completions (shim)  | ✅           |
| Streaming (SSE)          | ✅           |
| Vision (image input)     | ✅           |
| Tools / function calling | ✅ parallel  |
| JSON mode                | ✅           |
| Grounding (web search)   | ✅           |
| Long context             | ✅ 1M tokens |
| Context window           | 1M tokens   |

## When to use

* **Mid-tier reasoning at low cost** — harder tasks than Flash can handle, without stepping up to Pro pricing.
* **Long-context work** — 1M context for codebase reviews and multi-doc QA where 200K isn't enough.
* **Fresh-fact tasks** — grounding (Google Search) is supported.

When **not** to use:

* Cheapest possible chat — `gpt-5-4-mini` is lower cost for simpler tasks.
* Hardest reasoning — `claude-opus-4-8` is a flagship sized for it.

## Next

* [gpt-5-4-mini](/models/gpt-5-4-mini) — fast, low-cost OpenAI tier
* [claude-haiku-4-5](/models/claude-haiku-4-5) — fast Anthropic small tier