Gemini 2.5 Flash Lite - Starrise AI

Gemini 2.5 Flash Lite is the lightest and most efficient language model in Google’s 2.5 series, available through Starrise AI via the native Gemini API. Ideal for large-scale scenarios that are sensitive to response speed and cost.

Key Capabilities

Native Gemini API — Uses Google’s native API format for full feature access
Ultra-efficient — Lowest latency and lowest cost in the Gemini 2.5 series
Multi-turn conversation — Conversations with system instructions
High throughput — Suitable for large-scale, cost-sensitive workloads

Quick Example

curl "https://ai.alad.com/v1beta/models/gemini-2.5-flash-lite:generateContent?key=YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "contents": [
      {
        "role": "user",
        "parts": [{ "text": "Explain quantum computing in simple terms." }]
      }
    ],
    "generationConfig": {
      "temperature": 1,
      "topP": 1
    }
  }'

import requests

url = "https://ai.alad.com/v1beta/models/gemini-2.5-flash-lite:generateContent"
params = {"key": "YOUR_API_KEY"}
data = {
    "contents": [
        {
            "role": "user",
            "parts": [{"text": "Explain quantum computing in simple terms."}]
        }
    ],
    "generationConfig": {
        "temperature": 1,
        "topP": 1
    }
}

response = requests.post(url, params=params, json=data)
result = response.json()
print(result["candidates"][0]["content"]["parts"][0]["text"])

Parameters

Parameter	Type	Required	Description
`key`	string	Yes	API key (query parameter)
`contents`	array	Yes	Array of `{ role, parts }` objects
`systemInstruction`	object	No	System instruction with `parts` array
`generationConfig.temperature`	float	No	`0`–`2`, controls randomness, default `1`
`generationConfig.topP`	float	No	Nucleus sampling threshold, default `1`

API Reference

View the interactive API Playground for Gemini 2.5 Flash Lite.