Aion 3.5 Mini
GPT-6 Luna
- Context
- 1Mtokens
- Released
- 1y ago
Google DeepMind's Gemini 2.5 Flash-Lite is a new addition to the Gemini 2.5 model family, introduced in preview and described by the company as its fastest and most cost-efficient 2.5 model so far. Google says it scores higher than 2.0 Flash-Lite on coding, math, science, reasoning and multimodal benchmarks while running with lower latency. The model is built for high-volume, latency-sensitive work such as translation, classification and simple data extraction, where cost and speed are the primary constraints.
- Multimodal model
- Accepts images, audio, video, files
- 1M context
- Reasoning
- Tools
How it compares
Level
LightweightThe lab's smallest, cheapest and lowest-latency hosted line, built for high volume and simple tasks.
Lightweight models from other labs
Vs. your stack
Add the models you use to see how this one compares.
Sign inAdditional source evidence
- Source model
- google/gemini-2.5-flash-lite-preview-06-17
- Source family
- gemini-flash-lite
- Last updated
- 2025-06-17
- Knowledge cutoff
- 2025-01
- Model context
- 1M
- Model output limit
- 66K
- Open weights
- No
- File attachments
- Yes
- Reasoning
- Yes
- Tool calling
- Yes
- Temperature control
- Yes
Complete source record (JSON)
{
"id": "google/gemini-2.5-flash-lite-preview-06-17",
"name": "Gemini 2.5 Flash-Lite Preview 06-17",
"description": "First preview of Gemini 2.5 Flash-Lite, the low-cost, low-latency 2.5 model",
"family": "gemini-flash-lite",
"last_updated": "2025-06-17",
"knowledge": "2025-01",
"attachment": true,
"reasoning": true,
"tool_call": true,
"temperature": true,
"modalities": {
"input": [
"text",
"image",
"audio",
"video",
"pdf"
],
"output": [
"text"
]
},
"limit": {
"context": 1048576,
"output": 65536
},
"release_date": "2025-06-17",
"open_weights": false
}Text provenance
Gemini app/webGoogle SynthID text watermarking
Google DeepMind says text generated in the Gemini app and web experience is watermarked with SynthID by adjusting token probabilities.
Official explanationFrequently asked questions
Capabilities and specs
The context window is 1M tokens (1,048,576), and the maximum output is 66K tokens (65,536).
Yes. Its input modalities are text, image, audio, video and file.
Yes, tool calling (function calling) is supported.
Yes, a reasoning mode is supported.
The knowledge cutoff is 2025-01.
Quality and benchmarks
On Arena Vision it ranks 100 (Arena's own rank) with a score of 1189. On Arena Text it ranks 197 (Arena's own rank) with a score of 1375.
Matched on model kind and release date (no comparable price): Apple On-Device Model (2025) by Apple, Apple Server Model (2025) by Apple, and Qwen3 Coder Plus by Alibaba.
Apple On-Device Model (2025)Gemini 2.5 Flash-Lite Preview 06-17 vs Apple On-Device Model (2025)Apple Server Model (2025)Gemini 2.5 Flash-Lite Preview 06-17 vs Apple Server Model (2025)Qwen3 Coder PlusGemini 2.5 Flash-Lite Preview 06-17 vs Qwen3 Coder Plus
Gemini 2.5 Flash-Lite Preview 06-17 is closed, so it has no open-weight counterpart. The highest-ranked open-weight model on Arena Text is Kimi K3 at rank 19.
Versions and lineage
Both models have a context window of 1M tokens (1,048,576) and were released on June 17, 2025. Gemini 2.5 Flash is priced at $0.30 input and $2.50 output per 1M tokens.
Gemini 2.5 FlashGemini 2.5 Flash-Lite Preview 06-17 vs Gemini 2.5 Flash
Licensing and access
No, Gemini 2.5 Flash-Lite Preview 06-17 is closed.
Basics
Gemini 2.5 Flash-Lite Preview 06-17 was made by Google DeepMind and released on June 17, 2025.
Generated from Vertical Tabs data, updated November 10, 2026.
Discussion
No comments yet
Be the first to share your thoughts.