Gemini 3.1 Pro

Gemini 3.1 Pro (google/gemini-3.1-pro-preview) is Google’s flagship model. Besides text it accepts images, documents, video and audio, and regulates its reasoning depth across five levels. Requests go through the usual POST /v1/chat/completions in OpenAI format.

Parameters and capabilities

Capability	Value
model	google/gemini-3.1-pro-preview (or the short gemini-3.1-pro-preview)
Image input	Yes — image_url with a data: URL in the last message
File input	Yes — PDF, DOCX, TXT, CSV, XLSX, PPTX, up to 5 per request
Video input	Yes — the capabilities.video flag: attachments in the web cabinet and the bot
Audio input	Yes — the capabilities.audio flag: attachments in the web cabinet and the bot
Reasoning	reasoning_effort: low, medium, high, xhigh, max — cannot be disabled
Cache read	$0.25 per 1M tokens
Cache write	$0.47 per 1M tokens

The model’s reasoning cannot be turned off (can_disable_reasoning: false): it thinks on every request, and that is reflected in the price. Reasoning is billed as output tokens and eats into the max_tokens budget. Over /v1, images and documents work as attachments; video and audio input are attachments of the web cabinet and the bot.

Examples

A request with a reasoning depth

Python

from openai import OpenAI

client = OpenAI(base_url="https://api.mixen.ai/v1", api_key="mxn-...")

resp = client.chat.completions.create(
    model="google/gemini-3.1-pro-preview",
    messages=[{"role": "user", "content": "Compare these two pricing plans and find hidden contradictions"}],
    reasoning_effort="high",
)
print(resp.choices[0].message.content)

cURL

curl https://api.mixen.ai/v1/chat/completions \
  -H "Authorization: Bearer $MIXEN_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "google/gemini-3.1-pro-preview",
    "messages": [{"role": "user", "content": "Compare these two pricing plans and find hidden contradictions"}],
    "reasoning_effort": "high"
  }'

Streaming in Python

stream = client.chat.completions.create(
    model="google/gemini-3.1-pro-preview",
    messages=[{"role": "user", "content": "Walk through this contract point by point: risks, terms, penalties"}],
    stream=True,
    stream_options={"include_usage": True},
)
for chunk in stream:
    if chunk.choices and chunk.choices[0].delta.content:
        print(chunk.choices[0].delta.content, end="", flush=True)

Price

Per 1M tokens: input — $2.50, output — $15. Context cache: read — $0.25, write — $0.47. Catalog prices track the exchange rate — take current values from the catalog.

Strengths and limits

• Four kinds of input in one model: images, documents, video and audio — per the catalog flags.
• Five reasoning depth levels, including xhigh and max — from quick answers to multi-step analysis.
• A cache friendlier than most flagships: the $0.47 write is cheaper than the $2.50 full input, and reads cost roughly a tenth.
• Reasoning cannot be disabled — the lowest depth is low, so the model thinks even on simple questions.
• Output at $15 per 1M tokens is six times the input; long answers are billed in full.
• The model id carries a -preview suffix — pass it exactly as in the catalog.

All models — in the catalog. The general text workflow — chat guide.