Google said to launch Gemini 3.8 Live: real-time voice with synchronized multi-step reasoning
xiaohu · x · 2026-09-16
A blogger claims Google released Gemini 3.8 Live plus an Extended Thinking variant, pushing real-time voice into multi-step reasoning (unverified — await official confirmation).
- Base version: optimized for high-concurrency, low-latency, cost-effective deployment; automatic language detection and smooth switching across 97 languages; real-time multimodal (vision + voice) interaction for assistants and retrieval services.
- Extended Thinking: enterprise-grade multi-step reasoning and planning for customer service, troubleshooting, and multi-system coordination.
- Standout feature: thinking and speaking run in sync — the model naturally says "let me check…" and reports progress while making async tool calls (external APIs, retrieval) in the background.
More from Models
- Fable 5.5 creates 66-second animated short entirely with self-written code, no image models — oran_ge · 2026-11-02
- SenseTime open-sources SenseNova-Loop-9B: weight-shared 56-layer loop, 256K context, Apache 2.0 — AdinaYakup · 2026-11-02
- User says Gemini excels at long-form fanfic, producing stories up to 29 minutes of reading — Ok-Jello3739 · 2026-11-02
- Sam Altman calls tokens 'a kind of fake unit' that doesn't measure business value — rohanpaul_ai · 2026-11-02
- 'You Can Feel the Gradient Slopes': Dev Contrasts Benchmaxxed RL Models vs General Ones — theteknosaur · 2026-11-02
- LightOnOCR-3 Launches: Tops FRBench, Adds Layout Structure and Chart Extraction — IgorCarron · 2026-11-02