Lifestyle
Google launches Gemini 3.8 Live with Live Avatar: a talking, expressive AI avatar for businesses
On September 24, 2026, Google announced Gemini 3.8 Live with Live Avatar, which adds near-real-time generated avatar video to live voice conversations, available first in Gemini Enterprise. This article summarizes the features, limitations and what it means for ordinary users, based on official Google materials.
About 5 min read
What Google announced
On September 24, 2026, Google published Gemini 3.8 Live with Live Avatar on its official blog. Google said this is a new feature following last week's release of Gemini 3.8 Live, combining near-real-time video generation with voice so the AI can converse with people as an animated avatar that listens, watches and speaks.
According to Google, Live Avatar offers precise lip sync, natural facial expressions and smooth turn-taking, and can be used in business scenarios such as customer service or interactive guides. Google said the feature is available in Gemini Enterprise from launch day, with API documentation for developers (an API is the interface developers use to connect their own software to Google's AI).
Read the full description
Sources are collected, independently checked, then reviewed by Jev.
Key features: what Google says
- Multimodal input (handling several kinds of input at once): Google says Live Avatar processes visual and audio input simultaneously, then responds with expressive voice and video.
- Asynchronous tool calling (doing lookups in the background without pausing the conversation): Google says the avatar can call tools and retrieve data while the conversation continues; the demo scenario is a hotel guest check-in.
- Multilingual: Google says it can switch among 97 languages and claims this does not reduce video fidelity or cause visual drift, meaning the avatar's appearance does not gradually degrade or shift.
- Custom avatars: beyond the preset avatar library, developers can generate custom avatars from high-quality reference images; Google says this is currently available only through an enterprise allowlist, meaning only business customers Google has approved.
Gemini 3.8 Live vs. Live Avatar
| Item | Gemini 3.8 Live | With Live Avatar |
|---|---|---|
| Output types | Audio and text | Audio, video and text |
| Output limit | 64K tokens | 24K tokens |
| Availability | Gemini API, Gemini App, Google AI Studio, Gemini Enterprise Agent Platform, Google Search Live (per model card) | Gemini Enterprise (per Google blog) |
| Continuous interaction time | Not separately stated in model card | Minutes, not hours (per model card) |
According to the Google DeepMind model card, Gemini 3.8 Audio is based on Gemini 3 Pro, and Gemini 3.8 Live's input can include audio, images, video and text, with a context window (the amount of material the model can take in at once) of up to 128K tokens. The model card also mentions general foundation-model limitations such as possible hallucinations (confidently stating things that are not true), occasional slowness or timeouts, and a knowledge cutoff of January 2025.
Safety and transparency: SynthID watermarks
Google says all outputs generated by its AI products carry a SynthID watermark embedded directly in the audio and video. According to Google DeepMind, SynthID watermarks are imperceptible to humans but detectable by SynthID technology; users can upload an image, video or audio to Gemini and ask whether it was generated or modified by Google AI. Google DeepMind also says the SynthID Detector verification portal is currently being tested with journalists and media professionals.
The model card further states that, based on evaluation results for Gemini 3.7 Flash, Google considers models such as Gemini 3.8 Live unlikely to reach the tracked or critical capability levels in its Frontier Safety Framework, Google's own system for flagging potentially dangerous AI abilities. This is Google's own assessment, not the conclusion of an independent external review.
FAQ
Can ordinary people use Live Avatar now?
According to the Google blog, Gemini 3.8 Live with Live Avatar is currently offered in Gemini Enterprise, mainly for businesses and developers; custom avatars are further restricted to an enterprise allowlist. Ordinary users are more likely to encounter it indirectly through a company's customer service or guide services.
What languages can Live Avatar speak?
Google says Live Avatar can switch among 97 languages and adjusts lip sync and expressions when switching languages mid-conversation. Google's announcement does not list the full set of languages.
Can I talk with it for a long time?
No. The Google DeepMind model card states that the version with Live Avatar supports minutes, not hours, of continuous interaction; it may also occasionally be slow or time out.
How can I tell the person on screen is AI-generated?
Google says outputs from its AI products carry an embedded SynthID watermark. According to Google DeepMind, you can upload an image, video or audio to Gemini and ask whether it was generated or modified by Google AI. This method only identifies Google AI content.
Have these feature claims been independently verified?
No. All information in this article comes from official Google and Google DeepMind pages, including the features, the number of languages and the safety assessment; all are Google's own claims.
Browse the latest news in this topic
Lifestyle
NVIDIA launches DGX Spark 64GB: on sale October 23 from $4,999, two units can be linked into 128GB
On October 2, 2026, NVIDIA announced a more affordable 64GB memory version of its DGX Spark personal AI computer, available from October 23 through six makers including Acer and ASUS. It is aimed mainly at developers and researchers who want to run AI models on their own machines. Below we summarize the specs NVIDIA published, its claims about linking two units, and what it means for general readers.
Lifestyle
Google Cloud Launches Spanner Queues: Putting Message Queues Inside Database Transactions to Make AI Agents More Reliable
Google Cloud has announced the general availability of Spanner queues, which make message creation part of a database transaction. The aim is to stop AI agents' "state" and "actions" from falling out of sync. This article covers Google Cloud's claims, the main features, and what it means for general readers.
Lifestyle
GPT-6.1 Sol Launches: New Sol Version in the API, Codex and ChatGPT Work, Not in Chat
OpenAI launched GPT-6.1 Sol on September 29, 2026, with the API name gpt-6.1-sol. The launch rollout covers Codex and ChatGPT Work on Plus, Pro, Business, Enterprise and Edu (Enterprise and Edu need an administrator to enable it); Free and Go are not included at launch, and it is not in Chat (checked September 2026).
Lifestyle
Claude Sonnet 5.5 Launches: Same List Price as Sonnet 5, Available in the API, on Cloud Platforms and in Claude.ai
Anthropic launched Claude Sonnet 5.5 on September 28, 2026. API list prices are the same as Sonnet 5 ($2 per million input tokens, $10 per million output tokens). It is available in Claude.ai, the API and several cloud platforms, and higher-risk cybersecurity requests fall back to Sonnet 5 (checked September 2026).
Latest travel guides
GuideTokyo
Where to Stay in Tokyo: Comparing Shinjuku, Ueno, Tokyo Station, Shibuya, Asakusa, Ikebukuro, and Ginza, Plus Airport Access, Accommodation Tax, and Luggage Delivery
Where should you stay in Tokyo? Compare Shinjuku, Ueno, Tokyo Station, Shibuya, Asakusa, Ikebukuro, and Ginza by the same criteria: access from Narita and Haneda, transit routes, nearby attractions, neighborhood character, and who each area suits. Includes a comparison table, a Yamanote Line diagram, Tokyo’s accommodation tax as verified in 2026/9 (changing to 3% in 2027/4), and Airport TA-Q-BIN luggage shipping rules.
- Budget
- Hotels
GuideTokyo
How to Choose Tokyo Transit Passes: Are Suica, Welcome Suica, the Tokyo Subway Ticket, and the JR Pass Worth It?
On a first Tokyo trip, start with an IC card and pay per ride (Welcome Suica has no deposit and is valid for 28 days). If you take four or more subway rides in a day, add a 72-hour Tokyo Subway Ticket for 2,000 yen; a JR Pass is never worthwhile if you stay in Tokyo and do not go to Kansai. See what TOURIST PASMO, Suica on iPhone, and the Tokyo Metro day pass do and do not cover, with a decision chart. Prices verified in September 2026.
- Transport
- Budget
GuideTokyo
Tokyo Disneyland and DisneySea Guide: Ticket Prices, Fantasy Springs, Disney Premier Access (DPA), Standby Pass, and Which Park to Choose for Your First Visit
Tokyo Disney one-day Passport prices vary: most weekdays in 9/2026 cost ¥9,900 and weekends ¥10,900. At 14:00 daily, tickets go on sale for the same date two months later. Free Priority Pass is no longer on the official service list; only paid Disney Premier Access (¥1,000–3,500 per person per use) shortens waits. Covers hours, the 25th anniversary, Standby Pass, Entry Request, Fantasy Springs access and first-visit park choice; checked on the official site in 9/2026.
- Itineraries
- Family
Sources
- Introducing Gemini 3.8 Live with Live Avatar · Checked:
- SynthID — Google DeepMind · Checked:
- Gemini 3.8 Audio - Model Card — Google DeepMind · Checked: