Gemini 3.8 Flash

Google DeepMind · released 2026-09-02 · proprietary (API-only) · announcement

Price

Input $0.75/1M

Cached read $0.075/1M

Output $3.75/1M

$ / 1M tokens, primary tier. full pricing reference

unknown — not stated on announcement page; ai.google.dev models list confirms the endpoint ID only

not published as a table by Google DeepMind.

no retirement announced.

no merged support PR yet on the runtimes AI Squawk watches (vLLM, llama.cpp, transformers, SGLang, Ollama), and not on the Ollama library.

up no open incidents at Google DeepMind · full scoreboard

TOOLING crewAI 1.15.23 — Add support for native Gemini 3.8 Flash Implement evaluation of the last traced run through AMP in crewai eval… · 46d ago
DOCS Google Gemini Model Deprecations changed — `gemini-3.8-flash-tts` · September 22, 2026 · No shutdown date announced · `gemini-3.8-flash-lite-tts` · September 22, 2026 · No … · 51d ago
OUTAGE GitHub: Degradation with Gemini 3.8 Flash — On September 16, 2026, between 04:40 and 11:45 UTC, the Gemini 3.8 Flash model in GitHub Copilot experienced degraded availability. · 59d ago
TOOLING v2.21.0 — Add Gemini 3.8 Flash model to SDKs and update Flash model descriptions (4a969fe) · 72d ago
RELEASE Introducing Gemini 3.8 Flash and 3.8 Flash Cyber — · 72d ago
TOOLING Setup OpenClaw 2.0 with Gemini in Under 60 Seconds — Five commands, zero fluff: install OpenClaw 2.0, connect Gemini 3.8 Flash, and start chatting with live Google Search grounding from your… · 75d ago