Gemini 3.5 Flash
google/gemini-3.5-flash-20260519
Google's Gemini 3.5 Flash is a multimodal model from Google that accepts text, image, video, file, and audio inputs and produces text output. It supports tool usage and reasoning, provides a context length of 1,048,576 tokens, and is priced at $1.5 per 1 M input tokens and $9 per 1 M output tokens. The model is not open‑weight and was released on 2026‑05‑19.
- tool use
- structured outputs
- reasoning
At a glance
providergoogle
releasedMay 19, 2026
context1.0M
input / M$1.50
Specifications
What the API gives you
- Context window
- 1.0M tokens
- Max output
- 66K tokens
- Input price
- $1.50 / 1M tokens
- Output price
- $9 / 1M tokens
- Cache read
- $0.15 / 1M tokens
- Modalities
- text, image, video, file, audio → text
- Knowledge cutoff
- Jan 1, 2025
- Released
- May 19, 2026
Recommendation coverage
Where this model appears
- #1Best LLM for Everyday Translation
- #2Best LLM for Translation
- #4Best LLM for Image Understanding
- #4Best LLM for OCR & Documents
- #4Best LLM for Vision & Documents
- #4Best LLM for Charts & Diagrams
- #4Best LLM for Document Parsing
- #4Best LLM for Roleplay & Character
- #5Best LLM for Competition Math
- #5Best LLM for Literary Translation
- #5Best LLM for Document Summarization
- #5Best LLM for Tool & Function Calling
- #5Best LLM for Marketing Copy
- #5Best LLM for a General Assistant
- #5Best LLM for Grounded RAG
Measured evidence
Public benchmark record
- LMArena Agent
- -1.12026-07-13
- LMArena Agent
- -0.72026-07-20
- LMArena Text
- 1480 Elo2026-07-16
- LMArena Text
- 1490 Elo2026-07-20
- LMArena Vision
- 1309 Elo2026-07-12
- LMArena WebDev
- 1493 Elo2026-07-16
- LMArena WebDev
- 1488 Elo2026-07-20
LMArena data licensed CC-BY-4.0. Aider (Apache-2.0) and SWE-bench (MIT) are open.
Catalog data via OpenRouter and models.dev. Last checked within the past 6 hours.