Gemini 3.8 Flash API: OneInfer and Google Endpoints

Gemini 3.8 Flash is served on OneInfer under the model identifier google/gemini-3.8-flash. Use the OneInfer cURL or Python quickstart below for an OpenAI-compatible call, or hit Google AI Studio directly with the native Gemini API.

Last verified 3 September 2026.

Use Gemini 3.8 Flash in 4 steps

Step 1 of 4

Try Gemini 3.8 Flash in the OneInfer playground without writing code, or create an API key in the Console to call it from your app.

export ONEINFER_API_KEY="your-api-key"

OneInfer cURL quickstart

Example
curl https://api.oneinfer.ai/v1/ula/chat/completions \
  -H "Authorization: Bearer $JWT_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "google/gemini-3.8-flash",
    "messages": [{"role": "user", "content": "Summarise the attached research in five bullets."}],
    "max_tokens": 2048,
    "stream": false
  }'

OneInfer Python quickstart

Example
from openai import OpenAI

client = OpenAI(
    base_url="https://api.oneinfer.ai/v1/ula",
    api_key="$JWT_TOKEN",
)

response = client.chat.completions.create(
    model="google/gemini-3.8-flash",
    messages=[
        {"role": "user", "content": "Summarise the attached research in five bullets."},
    ],
    max_tokens=2048,
)
print(response.choices[0].message.content)

Verified API facts

FieldValue
OneInfer model identifiergoogle/gemini-3.8-flash
Upstream model identifiergemini-3.8-flash
Input context1,048,576 tokens (1M)
Maximum output65,536 tokens (64K)
Input modalitiesText, image, audio, video, PDF
Output modalityText
Thinking levelslow, medium (default), high
Minimal thinking levelNot supported on this model
ToolsFunction calling, structured output, caching, code execution, file search, URL context, Google Search grounding, Google Maps grounding, computer use (preview)

Ready to test the workflow?

OneInfer availability

Gemini 3.8 Flash is live on OneInfer under google/gemini-3.8-flash. Point the OpenAI-compatible base URL at https://api.oneinfer.ai/v1/ula and pass your OneInfer API key in the Authorization header. The request body is unchanged from any other chat-completions call; only the base URL and the model string differ.

Day-zero third-party support

LiteLLM (docs.litellm.ai/blog/gemini_3_8_flash), Vercel AI Gateway, AIHubMix and LLM Gateway all published Gemini 3.8 Flash support on 2 September 2026. LiteLLM and Vercel both hold first-page Google positions with their changelog posts. If you need an OpenAI-compatible endpoint today, OneInfer is a working route.

Frequently asked questions

Can I run Gemini 3.8 Flash on OneInfer?

Yes. Gemini 3.8 Flash is served on OneInfer under the model identifier google/gemini-3.8-flash. Point the OpenAI-compatible base URL at https://api.oneinfer.ai/v1/ula and pass your OneInfer API key in the Authorization header.

What is the Gemini 3.8 Flash model ID on OneInfer?

Use google/gemini-3.8-flash as the model identifier against the OneInfer OpenAI-compatible endpoint. OneInfer resolves that identifier to Google's model.

Does Gemini 3.8 Flash support thinking level minimal?

No. Low, medium (default) and high are the three supported levels. Setting minimal returns an error.

How long does Gemini 3.8 Flash take to respond?

About 13.3 seconds to the first token, then about 305 tokens per second. Stream responses so the 13-second wait does not look like a hang.

Put Gemini 3.8 Flash to work

Fund a controlled evaluation, test it on your own inputs, and measure quality and cost on your workload.