Google's open multimodal embedding model that maps text, images, audio and video into one vector space, under Apache 2.0
Gemini 3.8 Live
New· 4Google's low-latency speech-to-speech model for voice agents that also takes video and image input
About Gemini 3.8 Live
Google's model for spoken conversation
Voice agents are the target of Gemini 3.8 Live, Google's real-time dialogue model, which began rolling out on 15 September 2026. Google describes the design goals as scale and low cost. The model accepts audio, images, video and text and answers in audio or text, and visual input is processed almost as it arrives, so an agent can talk about what a camera shows while the conversation goes on. It detects and switches between 97 supported languages in the middle of a conversation.
Technical profile
In the Gemini API the model is listed as a stable version under the identifier gemini-3.8-live. Input can run to 131,072 tokens, and answers to 65,536. Developers can use function calling, Search grounding and thinking, but not caching, structured outputs, code execution or batch processing. According to Google's model card, the Gemini 3.8 audio models build on Gemini 3 Pro, and every piece of audio they produce is marked with a SynthID watermark.
Where it is available
Developers get the model in the Gemini API and in Google AI Studio. It is in private preview in Gemini Enterprise, and consumers meet it in Search Live. The Gemini API pricing page puts it on the same price row as the Extended Thinking variant, so the two are billed alike.
How it ranks
Google reports that users ranked it second in the Speech Agent Arena. That is Google's own account of the result, and the result is cited here as a vendor claim.
The companion model
For tasks that need more reasoning during a conversation, Google offers a separate model, Gemini 3.8 Live Extended Thinking, which keeps the same input and output limits and adds deeper reasoning while it talks.
TrustList summary
Google started releasing Gemini 3.8 Live on 15 September 2026, a model for live spoken conversation. Google says it is built for scale and cost efficiency. It accepts audio, images, video and text and replies in audio or text, handling visual input in near real time during a conversation.
Main features
- Voice conversation: speech-to-speech dialogue designed for voice agents.
- Multilingual: detects and switches between 97 supported languages mid-conversation.
- Tool features: function calling, Search grounding and thinking are supported; caching, structured outputs, code execution and batch are not.
- Limits: it accepts 131,072 tokens in and returns up to 65,536 tokens out.
Who it serves
Developers through the Gemini API and Google AI Studio, Gemini Enterprise in private preview, and consumers through Search Live.
Key facts
- Model id: gemini-3.8-live, listed as stable
- Ranking: Google reports second place in the Speech Agent Arena, its own account
- Provenance: based on Gemini 3 Pro, with generated audio carrying a SynthID watermark
- Companion: a separate Extended Thinking model offers deeper reasoning
Summary written by TrustList from the business's own published information, Nov 2026.
Benchmarks & AI stats
| Benchmark | Official | Community avg |
|---|---|---|
| No benchmark scores yet. Be the first to add one. | ||
“Official” values are editor-approved and feed the ranking. “Community avg” is the mean of member submissions (shown for transparency; it never affects the ranking until an editor approves a value).
Sign in to add a benchmark score for this model.
Gemini 3.8 Live launches
Trust Score
An earned signal from verification, reviews, awards, transparency and engagement - the vendor can't buy it.
Updated 9/16/2026
Request a demo or quote from Gemini 3.8 Live
Reviews
Write the first review of Gemini 3.8 Live
Used it? Your experience helps other buyers decide.
Write a reviewQuestions & answers
No questions yet. Be the first to ask about Gemini 3.8 Live.
Gemini 3.8 Live alternatives
Similar listings buyers compare against this one.
Black Forest Labs' image generation and editing model with bounding-box layouts, multiple reference images and 4K output
Aleph Alpha's open-weight English-German mixture-of-experts model for self-hosting, released under Apache 2.0 in October 2026
Perplexity's open-weight decision model that picks among given options for routing and checks in AI systems
Open-weight translation model from Bilibili's Index team, covering 150 languages and following glossary and style instructions
Microsoft AI's text-to-speech model for keeping one voice across 23 languages, with voice cloning from short audio
Anthropic's smallest and fastest Claude 5.5 model, available on its own API and three cloud platforms
Liquid AI's open-weight decision model, giving yes or no, choice or score answers from text and images
More on TrustList
Everything here links back to the same verified catalogue. Pick your next stop.
- Write a reviewHelp other buyers decide
- CompaniesAgencies, consultancies and IT service providers, ranked by verified reviews.
- ProductsSoftware and SaaS with pricing, features, integrations and alternatives.
- ArticlesGuides, comparisons and research from the editorial desk and the community.
- AwardsAnnual recognition decided by verified reviews and an independent jury.
- LaunchesNew products and releases, voted up by the community every day.
- RequestsBuyers describe what they need; vendors respond directly.
- PeopleReviewers, authors and makers with public profiles.
- ComparePut up to four listings side by side before you shortlist.