Entity trail

Gemini 3.1 Flash Live

Source-backed findings, relationship evidence, citations, and briefing history from the public MindPattern archive.

Briefing refs

Findings

Edges

Sources

Corpus findings

2026-07-01 / news-researcherGoogle Ships Nano Banana 2 Lite and Gemini Omni Flash to DevelopersGoogle released Nano Banana 2 Lite (Gemini 3.1 Flash-Lite Image) generally — its fastest, cheapest image model, generating images in as little as four seconds at $0.034 per 1,000 images, live in AI Studio, the Gemini API, AI Mode in Search, and the Gemini app. Alongside it, Gemini Omni Flash entered public preview for video generation and conversational editing at $0.10 per second of output (matching Veo 3.1 Fast), currently capped at 10-second clips. The pairing pushes Google's cost-efficient multimodal stack directly into developers' hands.
2026-05-16 / github-pulse-researcherdograh-hq/dograh at 1.3K Stars With +297 Today: Open Source Voice Agent Platform Ships as Self-Hosted Vapi Alternative With Sub-300ms LatencyDograh is an open-source alternative to Vapi for building voice AI agents with a drag-and-drop workflow builder. It ships with built-in LLM/TTS/STT stack requiring no API keys, telephony integration (Twilio, Vonage), and true speech-to-speech via Gemini 3.1 Flash Live achieving sub-300ms latency with barge-in handling. BSD 2-Clause licensed, fully self-hostable, and supports local model deployment.
2026-05-09 / skill-finderGoogle Gemini 3.1 Flash-Lite Hits General Availability — $0.25/M Input Tokens, 2.5x Faster Than 2.5 Flash, Built for Agentic Pipelines at ScaleGoogle announced GA for Gemini 3.1 Flash-Lite on May 7, its fastest and most cost-efficient Gemini 3 model. Priced at $0.25/M input and $1.50/M output tokens, it delivers 2.5x faster time-to-first-token and 45% faster output generation versus 2.5 Flash with equivalent or better quality. Google positions it for tool calling, orchestration, and automated pipelines where cost-efficiency at volume matters. This is the cheapest frontier-adjacent model from a major lab.
2026-04-16 / news-researcherGoogle Launches Gemini 3.1 Flash TTS with 200+ Audio Tags and 70+ LanguagesGoogle released Gemini 3.1 Flash TTS, a next-generation text-to-speech model featuring 200+ audio tags for granular control over vocal style, pace, and delivery. The model supports 70+ languages, native multi-speaker dialogue without separate API calls, and achieved an Elo score of 1,211 on the Artificial Analysis TTS leaderboard. Available through Gemini API, Google AI Studio, Vertex AI, and Google Workspace Vids with SynthID watermarking.
2026-03-27 / sources-researcherGoogle Gemini 3.1 Flash Live: Real-Time Multimodal Voice Model — 90+ Languages, 90.8% Function-Calling AccuracyGoogle released Gemini 3.1 Flash Live, their highest-quality audio model for real-time dialogue. Supports 90+ languages, filters background noise, follows conversation threads 2x longer than previous model, scores 90.8% on ComplexFuncBench Audio for multi-step function calling with constraints. All audio is watermarked to prevent misinformation. Available via Gemini Live API in Google AI Studio and Search Live in 200+ countries. The 90.8% function-calling accuracy means agents can reliably trigger tool use mid-conversation.

Graph relationships

BENCHMARKED AGAINST
ComplexFuncBench Audio
Gemini 3.1 Flash Live scores 90.8% on ComplexFuncBench Audio.
Source finding
RELEASED
Google
Google released Gemini 3.1 Flash Live real-time audio model.
Source finding
USES
Dograh
Dograh achieves sub-300ms latency via Gemini 3.1 Flash Live
Source finding
RELEASED
Google -> Gemini 3.1 Flash Live
Google released Gemini 3.1 Flash Live real-time audio model.
Source finding
USES
Dograh -> Gemini 3.1 Flash Live
Dograh achieves sub-300ms latency via Gemini 3.1 Flash Live
Source finding

Source trail

Google DeepMind Blog GitHub Google Cloud Blog Google Blog Google DeepMind

Graph sources

entity graphfindings textkg edgeskg entities