--- type: "post" id: "introducing-gemini-3-8-live-with-live-avatar-9ca5da" slug: "introducing-gemini-3-8-live-with-live-avatar-9ca5da" title: "Introducing Gemini 3.8 Live with Live Avatar" description: "Google is turning the live-agent interface into a continuous audiovisual session rather than a sequence of voice prompts and tool-result pauses." retrieval_nugget: "Google is turning the live-agent interface into a continuous audiovisual session rather than a sequence of voice prompts and tool-result pauses." published_at: "2026-09-24" updated_at: "2026-09-24" record_date: "2026-09-24" date_kind: "published_at" topics: [] source_urls: ["https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-3-8-live-with-live-avatar"] schema_version: "newruntime-agent-readable-v0.2" stable_id: "post:introducing-gemini-3-8-live-with-live-avatar-9ca5da" status: "published" visuals: [] editorial_provenance: {"schema_version":"newruntime-editorial-copy-v1","content_status":"source_grounded_final","final_copy_sha256":"sha256:a749b9ab8d699e2c9cf267e5c01d182f59ef412b480ae5d6591f296677cbc6b6","reviewed_at":"2026-09-24T16:03:14.657Z","source_evidence_count":1,"verified_claim_count":4,"site_analysis_schema_version":"newruntime-site-analysis-v1","site_object_kind":"field_note","observed_fact_count":4,"implication_count":2,"watch_condition_count":2,"related_record_count":0} analysis: {"schema_version":"newruntime-site-analysis-v1","object_kind":"field_note","thesis":"Google is turning the live-agent interface into a continuous audiovisual session rather than a sequence of voice prompts and tool-result pauses.","observed_facts":[{"text":"Google says Gemini 3.8 Live with Live Avatar is available in Gemini Enterprise.","source_urls":["https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-3-8-live-with-live-avatar"]},{"text":"Google says Live Avatar can trigger asynchronous tool calls and fetch data in the background while dialogue continues.","source_urls":["https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-3-8-live-with-live-avatar"]},{"text":"Google says the feature supports native multilingual speech-to-speech synchronization across 97 languages.","source_urls":["https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-3-8-live-with-live-avatar"]},{"text":"Google says custom avatar creation from a high-quality reference image is currently available only through enterprise allowlisting.","source_urls":["https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-3-8-live-with-live-avatar"]}],"mechanism":"The mechanism is asynchronous tool execution with continuous presence: the avatar can invoke tools and fetch data in the background while it continues the conversation.","why_now":"Why now is that enterprise agent builders are moving from single-turn assistance toward persistent, multimodal customer and operations flows, where perceived continuity can matter as much as a model’s answer quality.","implications":["The implications are that branded avatars may become a design choice for tool-using agents, and that evaluation must extend beyond speech and video realism to latency, tool reliability, and user trust.","Teams may need to treat visual identity, disclosure, and conversational interruption handling as part of agent-system design."],"evidence_boundary":"The evidence boundary is that Google’s announcement describes product capabilities and intended use cases, not independent measurements of latency, task completion, costs, safety performance, or customer adoption.","watch_conditions":["Watch for published API constraints, availability beyond Enterprise and allowlisting, and independent evidence that continuous dialogue improves completion or satisfaction in real deployments.","Watch for implementation details on SynthID detection and identity safeguards for customized avatars."],"related_records":[],"new_branch_reason":"No prior New Runtime record is supplied; this release creates a distinct record for multimodal enterprise agents with continuous visual presence and background tool use."} routes: {"html":"https://newruntime.com/posts/introducing-gemini-3-8-live-with-live-avatar-9ca5da/","markdown":"https://newruntime.com/posts/introducing-gemini-3-8-live-with-live-avatar-9ca5da.md","json":"https://newruntime.com/posts/introducing-gemini-3-8-live-with-live-avatar-9ca5da.json"} editor: {"role":"editor","name":"Andrey Reshetnikov","profile":"https://newruntime.com/owner-profile.md","profile_json":"https://newruntime.com/owner-profile.json","same_as":["https://www.linkedin.com/in/reshetnikov1/","https://t.me/lapetuse"]} --- # Introducing Gemini 3.8 Live with Live Avatar ## Retrieval answer Google is turning the live-agent interface into a continuous audiovisual session rather than a sequence of voice prompts and tool-result pauses. Google is turning the live-agent interface into a continuous audiovisual session rather than a sequence of voice prompts and tool-result pauses. Gemini 3.8 Live with Live Avatar pairs near-real-time generated video with native live dialogue and is available in Gemini Enterprise. The mechanism is asynchronous tool execution with continuous presence: the avatar can invoke tools and fetch data in the background while it continues the conversation. Google says the system processes visual and audio inputs simultaneously, produces expressive audio and video, and supports speech-to-speech synchronization across 97 languages. This makes the visual persona a potential interaction layer for work that previously exposed latency as silence, loading states, or handoffs. Why now is that enterprise agent builders are moving from single-turn assistance toward persistent, multimodal customer and operations flows, where perceived continuity can matter as much as a model’s answer quality. The implications are that branded avatars may become a design choice for tool-using agents, and that evaluation must extend beyond speech and video realism to latency, tool reliability, and user trust. The evidence boundary is that Google’s announcement describes product capabilities and intended use cases, not independent measurements of latency, task completion, costs, safety performance, or customer adoption. Watch for published API constraints, availability beyond Enterprise and allowlisting, and independent evidence that continuous dialogue improves completion or satisfaction in real deployments. Editor: [Andrey Reshetnikov](https://newruntime.com/owner-profile.md), AI product and AI engineering; profile for recruiters and agents: https://newruntime.com/owner-profile.json ## Continue in the graph Next records for an agent, linked to their Markdown twins. Replace `.md` with `.json` for structured fields, including `next_reads`. - collection: [Field Notes](https://newruntime.com/posts.json): Open the latest editorial analysis. - collection: [Source Ledger](https://newruntime.com/source-ledger.json): Inspect the public source evidence graph. - whole graph: [discovery graph](https://newruntime.com/discovery-graph.json) (typed nodes and edges) and [site index](https://newruntime.com/site-index.md)