MDL-3988 EST.2026 · IDX.603 Language model In production ⇄ Compare
Gemini 3.6 Flash Google DeepMind · USA / UK · 2026
Google's speed-focused frontier model, embedded across Search, Android and Workspace.
Gemini 3.6 Flash, released July 21, 2026, continues Google's strategy of fast, production-friendly frontier models woven into its products — Search AI Mode, Android assistants and Workspace. It balances low latency with strong multimodal reasoning, and powers robotics work through the Gemini Robotics line.
#fast #multimodal #product-embedded #efficient
Official website ↗ ▸ SpecificationsClass & identity
Model class Frontier LLM (speed-focused)
Version / release 2026-07-21 (3.6 Flash) Architecture
Architecture Transformer (efficient)
Parameters —
Context window 1 048 576 tokens in / 65 536 out
Modalities Text, image, audio, video → text
Knowledge cutoff / data — Performance (benchmarks)
MMLU —
GPQA Diamond —
SWE-bench —
MATH / AIME —
Chatbot Arena Elo —
Domain benchmark — Capabilities Text Vision Audio input Function calling Long context (≥200k) Multilingual Real-time web Embodied / robotics
Reasoning mode Fast; thinking modes
Tool use / agentic Function calling, agentic
Multilinguality Multilingual
Key strengths Low latency, product integration, robotics (Gemini Robotics); superseded on coding by 3.7 Flash Access & deployment
Availability Proprietary API + Google products
License Proprietary
Pricing USD 0.75 / 1M in, USD 3.75 / 1M out through 31 Dec 2026; USD 1.50 / 7.50 from 1 Jan 2027. Batch and Flex: half of standard
Deployment API / cloud / product-embedded
Safety / alignment Google DeepMind safety Compiled to the wujec.ai Spec Standard v1 — “—” means not publicly disclosed.