OpenRouter launches Gemini 3.6 Flash and Flash-Lite

// 2h agoMODEL RELEASE

OpenRouter launches Gemini 3.6 Flash and Flash-Lite

OpenRouter has added Google's Gemini 3.6 Flash and Gemini 3.5 Flash-Lite to its platform. Gemini 3.6 Flash offers enhanced coding and planning with 17% fewer output tokens, while Flash-Lite delivers low-latency execution exceeding 150 tokens per second for subagents.

// ANALYSIS

Bringing Gemini 3.6 Flash and 3.5 Flash-Lite to OpenRouter significantly lowers the cost and latency barriers for building scalable multi-agent systems.

–High Throughput: Operating at 150+ tokens/second makes Flash-Lite highly effective for fast subagents and background document processing.
–Token Efficiency: Gemini 3.6 Flash requires 17% fewer output tokens, providing direct cost savings for developers.
–Agentic Focus: Both models emphasize enhanced planning and low-latency execution designed specifically for autonomous agent workflows.

// TAGS

openroutergoogle-geminigemini-3.6-flashgemini-3.5-flash-litellmai-agentsmodel-launch

DISCOVERED

2h ago

2026-07-21

PUBLISHED

2h ago

2026-07-21

RELEVANCE

8/ 10

AUTHOR

OpenRouter

// KEEP READING

More AI developer news from the feed

EXPLORE FULL FEED

OPEN SOURCE33m ago

ayghri/i-have-adhd skill enforces direct AI responses

i-have-adhd is an open-source skill designed by ayghri for AI coding assistants like Claude Code and OpenAI Codex. It embeds ten prompt rules into an agent's context to enforce concise, structured answers without conversational preamble.

OPEN SOURCE38m ago

Microsoft releases Ontology Playground for knowledge graphs

Microsoft Ontology Playground is an open-source web application designed to help developers and data architects learn about ontologies and Microsoft Fabric IQ. Built as a fully static, browser-based TypeScript tool with zero backend dependencies, it features an intuitive visual designer for constructing entity types and relationships, interactive graph exploration using Cytoscape.js, pre-built ontology catalogues, and seamless export capabilities to standard RDF/XML and JSON formats.

OPEN SOURCE38m ago

Outlines enforces structured LLM outputs via constrained generation

Outlines, developed by dottxt, is an open-source Python library that enforces strict structure on Large Language Model outputs during generation. Constraining token sampling at the logit level guarantees compliance with JSON schemas, regular expressions, or Pydantic models without brittle retry loops.