Talk by Thor Schaeff (Developer Experience, Google DeepMind) recorded at the AI Engineer World's Fair 2026, demonstrating live voice and media agents built with Gemini. Covers chaining image generation with the stateful Interactions API for an illustrated story, near real-time speech translation with Gemini 3.5 Live Translate, a voice agent that calls tools such as Lyria 3 for music generation, and how the Live API works (stateful WebSocket session, native audio model rather than a speech-to-text and text-to-speech pipeline, barge-in, tool calls, Google Search grounding, 90 languages). Ends with a voice and vision demo in Google AI Studio. The description states these as the talk's planned content; no comments or tags were supplied to independently verify the demos. # Channel Presentation Observations: Live Voice Agents with Gemini **Source:** Video `01a15ca6-a82e-70c0-a418-458b3b3b1a31`, titled "Can Your Agent Hear You Now? Building Live Voice Agents with Gemini — Thor Schaeff" **Evidence base:** Title and transcript only. No curated metadata was supplied, so no channel-level identity context is available. --- ## 1. Explicit Self-Description These are statements the speaker makes about himself and his role. They are creator claims, not independently verified facts. - **Role:** The speaker states, "my name is Thor. I work on the developer experience at Google DeepMind. So that mostly includes Gemini API and Google AI Studio." - **Colleague:** He introduces "my colleague Philip here as well, who's working on managed agents on the API." - **Event context:** He describes the session as "a bit of a prelook at the workshop we're doing tomorrow," and refers to "the largest AI engineer World's Fair in history." - **Self-referential title note:** He says, "Gemini made up this title," attributing the talk title to a model rather than to himself. --- ## 2. Intended Audience The transcript shows a live, in-person conference audience of developers, with direct address and calls to action. - **Developers as the primary audience.** The speaker repeatedly asks for feedback and usage reports, for example, "if you're using the Gemini models through the API, uh you know, let me know if you have any feedback." - **Audience participation.** The talk includes interactive segments such as the opening quiz: "Does anyone know what model that is?" and "Anyone has a story line they want to start with?" - **International audience acknowledged.** He notes, "we have a very very international audience and so I can speak in my preferred language," and later asks, "Who's listening in in uh Chinese?" - **Follow-up pointers to a workshop.** He directs attendees to "join me tomorrow morning in uh room 2009, I think." --- ## 3. Claimed Scope The speaker describes the scope as a demonstration-led overview of Gemini's multimodal and real-time capabilities, with a focus on the Live API. - **Stated focus:** "tomorrow we'll do kind of Gemini API especially focused on um live agents uh conversational agents real time uh conversational agents multimodal real-time conversational agents." - **Models and products named:** - Gemini TTS - Omni Flash (video generation and editing) - Nano Banana Light (image generation and editing) - Gemini 3.5 Live Translate - Gemini 3.1 Flash Live - Lyria 3 (music generation) - **Interfaces named:** The Interactions API, the Live API, Google AI Studio, and Google Search grounding. --- ## 4. Demonstrated Presentation Patterns These patterns are observed within this single video. They are not evidence of a channel-wide style. ### 4.1 Live Demonstrations as the Core Format Most segments are live demos, with the speaker narrating the mechanics as they run. - A live translation session with QR-code audience participation: > "I want you to scan uh this QR code on your phone." - A live multi-turn Gemini Live jukebox demo, with a music generation tool call. - A live AI Studio test of voice, vision, and Mandarin switching. ### 4.2 Code and Architecture Walkthroughs The speaker explains technical concepts in plain language, sometimes with code on screen. - Stateful chaining in the Interactions API: > "we're passing this previous interaction ID to the interactions API which means that we have the previous context of the previous interaction." - The distinction between native audio and a cascading pipeline: > "the model itself does not go through text, but rather the intelligence is baked into the audio model." ### 4.3 Product Promotion Framed Conversationally Product promotion is present, but it is delivered with informal asides. - Direct calls to try products: "do try that out." - A self-aware remark about promotion and cost: > "we spend a lot of money on this domain, AI.studio. So I do appreciate if you use that one. Um so we can get some ROI there." - An acknowledgment of limits in the demo's practical use: > "I know this application might not necessarily be that useful for your business context." ### 4.4 Informal, Improvisational Tone The transcript shows a conversational register with unscripted moments. - Audience improv for the story-canvas demo: "So anyone you know this is a little bit like improv theater now." - Acknowledged technical friction: "I'm having some trouble with my mouse here." - A mistake acknowledged on air: "I think my uh my Mandarin didn't work out so well there." ### 4.5 Pointers to Partners and Ecosystem - The speaker names third-party partners as ecosystem support: "LifeKit, Pipecat, they all actually here in the expo as well. I recommend you go visit them." --- ## 5. Limits of These Observations - **Single video only.** None of these observations establish a channel-wide pattern, recurring format, or consistent audience. - **No metadata.** Channel name, upload history, and affiliation beyond the speaker's own statements could not be verified. - **Motivations not inferred.** The speaker's reasons for promotion or for the informal tone are not stated, so none are attributed here. - **Transcript quality.** The transcript contains garbled or unclear passages, including the "Taipe" spelling, "Dankur," and a partly rendered demo exchange. Quoted wording may not match the audio exactly. # Delivery Observation — Thor Schaeff, "Can Your Agent Hear You Now?" ## Tone & Register - **Conversational and informal throughout.** Openings and transitions are casual: *"Can you hear me? Fantastic. Hello everyone. Well, hello couple people that wandered in."* - **Self-deprecating and light-hearted**, especially about logistics: *"I'm sorry the schedule has been changing around"* and *"I think my uh my Mandarin didn't work out so well there."* - **Enthusiasm expressed through repeated superlatives**, applied to nearly every product shown: *"a really incredible model for expressive speech,"* "a really incredible model" (live translate), *"it is a really really cool um way to interact."* ## Vocabulary & Recurring Phrases - **"really incredible"** recurs as a default praise phrase for Gemini TTS, live translate, and the interactions API. - **Frequent audience-directed invitations**: *"do try that out,"* *"So, do try that out,"* *"if you want to try it out... you can add it to your um applications now."* This repetition is a stable pattern. - **Habitual interjections** typical of spoken delivery: *"uh,"* *"you know,"* *"kind of"* (e.g., *"kind of the key features here,"* *"kind of on the Gen Media side"*). - **Feedback solicitation to attendees**: *"let me know if you have any feedback"* and *"let us know as well"* — he positions himself as a direct channel to the team. - **Humor from product self-awareness**: *"Gemini made up this title. So I uh can your agent hear me now? I think that that is the title."* ## Teaching Habits - **Audience participation openers.** He repeatedly polls the room: *"Does anyone know what model that is?... It starts with G."* and *"Has anyone tried out Gemini 3.5 live translate? Yes. Okay, a couple folks."* - **Improv-style co-creation**: *"anyone knows this is a little bit like improv theater now. Anyone has a story line they want to start with?"* — the demo input comes from an attendee. - **Live demos as the core teaching method**, chained back-to-back: *"We're we're just going live demo to live demo here."* - **Translating playful demos to business use.** After the DJ demo he immediately generalizes: *"I know this application might not necessarily be that useful for your business context... but I think the individual components are definitely useful. So if you're building real-time conversational agents... this could also just call out to your CRM."* - **Explain-then-concrete pattern**: conceptual description first ("stateful websocket API"), then where to try it ("ai.studiolife," "AI.dev"). - **Light humor about budget/ROI of domains**: *"We spend a lot of money on this domain, AI.studio. So I do appreciate if you use that one. Um so we can get some ROI there."* ## Communication Patterns - **Interactive verification during demos** — he checks with the audience in real time: *"Do you get some audio out? Can you hear it? Yeah. Does it sound good? Fantastic."* - **Direct model-to-audience dialogue** as demonstration (speaking English, Mandarin, German on stage). - **Structural signposting is loose**; transitions lean on "Okay, so" repeatedly: *"Okay, so this was you know a bunch of things that happened."* / *"Okay, so if you're now um you know, you're like, 'Okay, this sounds great.'"* ## Confidence Notes - The **"really incredible"** and **"do try that out"** repetitions are well-supported across the transcript and appear stable. - His repeated use of *"uh"* and *"kind of"* is verifiable, but attribution to a consistent speaking style vs. nerves in a live setting is uncertain from a single recording. - The opening audience quiz ("It starts with G") is observed once; whether quizzing is a habitual opener cannot be confirmed from this video alone. # Observation Record: "Can Your Agent Hear You Now? Building Live Voice Agents with Gemini" **Source:** Video `01a15ca6-a82e-70c0-a418-458b3b3b1a31` **Speaker of record:** Thor Schaeff (identified by curated title metadata and self-introduction in transcript) **Other named individuals:** Philip, a colleague mentioned but not heard speaking in the transcript. No positions are attributed to Philip. --- ## 1. Speaker Context - **Self-described role:** Works on developer experience at Google DeepMind, covering the **Gemini API** and **Google AI Studio**. - **Stated purpose of talk:** A preview of a workshop scheduled for the following day, focused on live, conversational, multimodal real-time agents. --- ## 2. Unattributed Audio - **Opening clip:** A synthesized voice plays a welcome message ("Welcome one and all to the grandest stage on Earth…"). - Thor says the clip is **Gemini TTS**, which he describes as a strong model for expressive speech. - The clip's voice is not attributed to Thor. Only his description of the model is recorded here. --- ## 3. Claims About Models and Products ### 3.1 Gemini TTS (Text-to-Speech) - **Claim:** "Actually a really incredible model for expressive speech." - **Stated features:** - Inline audio text steering - Speech streaming - Latency "a lot lower… than it used to be" - **Stated recommendation:** Good option for those "in the market" for text-to-speech. ### 3.2 Generative Media Models - **Omni Flash:** Described as a video generation and editing model, stated as shipped "this morning." - **Nano Banana Light:** - Described as "really exciting." - Claimed to cut image generation and editing latency "quite significantly." - The slide deck was described as already out of date because of this release. - **Image generation timing:** Thor said image generation "still takes… two to three seconds," which he called "pretty impressively fast." ### 3.3 Interactions API - **Claimed nature:** A **stateful** API that allows interactions to be chained. - **Demonstrated mechanism:** Passing a `previous interaction ID` supplies prior context. - **Stated benefit:** Maintains **visual and character consistency** across generated images. - **Stated scope:** "Our unified API for interacting with all of the Google DeepMind models as well as agents." ### 3.4 Gemini 3.5 Live Translate - **Claims:** - Near-real-time translation. - Automatically detects the speaker's language. - Translates "as soon as the model is recognizing kind of the context." - "Really incredible." - **Availability (stated):** Available on the Gemini API for developers to add to applications. - **Demo conditions:** A live session was run with audience members scanning a QR code and selecting a language. Thor asked whether the audio was audible and whether it "sound[ed] good," and the audience response is not captured in the transcript. ### 3.5 Gemini 3.1 Flash Live - **Claims:** - Described as the "most recent real time conversational model." - Stated to be the model used in the Gemini app when speaking to Gemini. - **Live API characteristics (as described):** - Low-latency, real-time voice and vision API. - A **stateful WebSocket** session that accepts text, audio, and video frames and returns audio chunks plus transcripts. - A **native audio model**, explicitly contrasted with "cascading voice agent pipeline[s]": > "the model itself does not go through text, but rather the intelligence is baked into the audio model." - Supports **90 languages** in preview, enabling fluid language switching. - Supports **barge-in** (interrupting the model). - Supports tool calls. - Provides audio transcription. ### 3.6 Grounding with Google Search - **Claim:** The model "does not have access to real time information" on its own. - **Proposed remedy:** Grounding with Google Search, described as "basically like a built-in function call." - **Demo examples given:** Today's weather and why Germany played well in the World Cup. These were suggested queries, not reported results. --- ## 4. Demonstrations and Their Stated Purposes | Demo | Stated Purpose | Thor's Qualification | |---|---|---| | **Story canvas** (Gemini Live + Nano Banana Light) | Illustrate audience-prompted story with chained image generation | Described as "a little bit like improv theater"; the mouse was not working well | | **Live translate** | Show near-real-time translation to audience language | Stated that some audience members may need headphones | | **AI radio jukebox** (Gemini Live + Lyria 3 via function call) | Show real-time voice agent with tool calls to a music model | "I know this application might not necessarily be that useful for your business context," but the "individual components are definitely useful" | | **AI Studio live demo** (vision, Mandarin, grounding) | Show voice, vision, and language switching in AI Studio | Mandarin attempt described as not having "work[ed] out so well"; the response to a question about what the model ate was characterized as "it doesn't need to eat because it's a model" | **Demo outputs** (e.g., the model describing a black shirt and hat, or the jukebox lyrics) are recorded as model outputs shown during demos. They are not verified claims by the speaker. --- ## 5. Stated Recommendations and Calls to Action - Use **Google AI Studio** to try the Gemini Live model and understand its capabilities. - Fork the jukebox demo in AI Studio. - Consider the Live API for real-time agents that need multilinguality, capabilities, or tool calls. - Try the **Interactions API**. - Add **Gemini Live Translate** to applications and share what is built with it. - Visit partners **LiveKit** and **Pipecat** in the expo, which the speaker described as working on real-time media. - Attend the workshop the following morning (room number stated as "2009, I think," with the speaker noting the schedule may change). **Stated rationale for the domain:** The speaker said Google spends "a lot of money on this domain" (ai.studio) and asked the audience to use it "so we can get some ROI there." Recorded as a stated preference, not a verified fact. --- ## 6. Qualifications and Tensions - **Latency claims are relative and uncited.** "Quite significantly" and "a lot lower" have no numbers given in the transcript, except the 2–3 second image generation figure. - **Capabilities are presented as demonstrations, not benchmarks.** The speaker did not provide accuracy or reliability measures. - **Scaling caveat:** Real-time media "can be quite challenging to scale," which is presented as a reason for partner solutions. - **Demo reliability:** The speaker acknowledged hardware trouble (mouse) and a failed language attempt (Mandarin). Live output was not guaranteed. - **Business applicability:** The jukebox demo was explicitly described as possibly not useful for business contexts, while its components were described as useful. - **Humor and self-disclaimers:** The speaker said "Gemini made up this title," and the title is therefore not presented as an authored claim. --- ## 7. Transcript Artifacts (Unresolved) The transcript appears to contain automated-transcription errors. These are noted, not corrected: - "Gemini Life" likely corresponds to "Gemini Live" (appears many times, including "Gemini Life is getting a tool call," "Gemini 3.1 Flash Life"). - "LIA 3" likely corresponds to "Lyria 3" (music generation model named in the jukebox demo). - "Taipe's" and "Dankur" are probable transcription errors. - The model name rendered as "Gemini 3.5 live translate" is stated as is, with no independent confirmation of the version number. Whether these are transcription errors or actual product names should be verified against primary sources before citation. --- ## 8. Self-Check - [x] Every substantive claim is attributed to Thor Schaeff, the speaker of record. - [x] Model outputs, demo audio, and the unattributed opening clip are labeled as such. - [x] Qualifications, demo caveats, and the business-applicability disclaimer are preserved. - [x] No private motivations or beliefs are inferred beyond stated rationale. - [x] Transcript uncertainties are flagged rather than silently corrected. # Speaker Attribution & Provenance Record ## Video Identification - **Title:** *Can Your Agent Hear You Now? Building Live Voice Agents with Gemini — Thor Schaeff* - **Video ID:** `01a15ca6-a82e-70c0-a418-458b3b3b1a31` ## Speakers & Roles | Speaker | Role | Basis | |---|---|---| | **Thor Schaeff** | Presenter / Google DeepMind developer experience engineer | Self-introduction: *"my name is Thor. I work on the developer experience at Google DeepMind"* covering Gemini API and Google AI Studio | | **Philip** | Colleague (audience-visible attendee) | Introduced as working on **managed agents on the API**; no recorded dialogue attributed to him | | **Gemini TTS (pre-recorded)** | Model voice playback | Opening welcome audio; presenter confirms it is "Gemini TTS" | | **Gemini Life / Live model (demo outputs)** | Model voice — radio DJ demo and AI Studio live demo | Live model responses transcribed inline (`>>` segments); explicitly demo content, not a human guest | | **Audience members** | Unnamed attendees | Brief audience answers (e.g., guessing the model, language preferences); not individually identified | ## Referenced Products, Models & Sources (as claimed by presenter) - **Gemini TTS** — text-to-speech model, described as expressive, streamable, lower latency - **Gemini Life (Gemini 3.1 Flash Live)** — real-time conversational model; stated as the model behind the Gemini app voice - **Omni Flash** — video generation/editing model; stated as shipped "this morning" - **Nano Banana Light** — image generation/editing model; stated as shipped "this morning" - **Gemini 3.5 Live Translate** — near-real-time live translation model on the Gemini API - **LIA 3 (Lyria 3)** — music generation model, invoked via function call in the DJ demo - **Interactions API** — described as Google DeepMind's unified stateful API for models and agents - **Live API** — stateful WebSocket API for real-time text/audio/video; 90 languages supported (stated as preview) - **Google AI Studio** — demo environment (`ai.studio/live`, `ai.dev/live`); demos stated as forkable - **Partner ecosystem** — LiveKit, Pipecat named as partners present at the expo ## Third-Party / Attendee Claims - Audience members confirmed hearing the live translation demo in German and Chinese (verbal confirmations only). - The Demos (story canvas, live jukebox) were stated as **vibe-coded** by the presenter / built by Gemini in Google AI Studio — provenance of demo code per presenter's own statements. ## Promotional Segments (flagged) - **Google product promotion throughout:** Gemini TTS, Nano Banana Light, Omni Flash, Live Translate, Live API, Interactions API, and Google AI Studio are promoted with direct calls to action ("do try that out", fork the demo, visit partner booths). - **Upstream promotion:** Repeated invitation to the presenter's **workshop tomorrow (room 2009, schedule uncertain per speaker)** — this session is described as a preview of that workshop. - **Domain plug:** Presenter jokingly notes money spent on the AI Studio domain and asks for ROI/feedback. ## Availability & Access Limits - Live API features described as **in preview**; 90-language support stated as preview capability. - Session schedule instability acknowledged; workshop room number uncertain ("I think... check the schedule"). - No paywall or recording-availability limits stated in the transcript. ## Uncertainty Notes - Model naming follows the transcript's phonetic spellings ("Gemini Life", "Nano Banana", "LIA 3", "Pipecat"); exact official product spellings (e.g., *Live* vs. *Life*, *Lyria*) are inferred, not confirmed in the source. - No curated metadata beyond the transcript and title was supplied; speaker identities beyond Thor and Philip are unnamed. Can you hear me? Fantastic. Hello everyone. Well, hello couple people that wandered in. Um, I'm sorry the schedule has been changing around. I'm sorry you didn't find the quiet working area you were looking for. All right. Um, we are looking at some agents with Gemini API. Now, before we kick off, let's have a little listen to this. Welcome one and all to the grandest stage on Earth. You're standing right at the heart of the largest AI engineer World's Fair in history. Let's build the future together, shall we? Does anyone know what model that is? Yeah. You you you you're not allowed. Anyone else? Anyone know which model? No. Any guesses? It starts with G. Gemini. Yes, it is a Gemini model. But which one? No, it's not life. That's incorrect. Yes. This is Gemini TTS. Is that what you were about to say? No. Okay. Um, so Gemini text to speech. Uh, actually a really incredible model for expressive speech. You have inline audio text you can steer it with. Um, you can stream the speech. It's a lot lower latency now than it used to be. So uh if you are in the market for some speech uh text to speech that is Gemini TTS is a good model. Now let's have a look at uh some of the things we're doing today. Uh first of all my name is Thor. I work on the developer experience at Google DeepMind. So that mostly includes Gemini API and Google AI Studio. So if you're using the Gemini models through the API, uh you know, let me know if you have any feedback. I have my colleague Philip here as well, uh who's working on managed agents on the API. So if you're using that, uh let us know as well. So this is a bit of a a prelook at the workshop we're doing tomorrow. Um so tomorrow we'll do kind of Gemini API especially focused on um live agents uh conversational agents real time uh conversational agents multimodal real-time conversational agents we're just packing everything in there um and so what we can look at actually there's uh been a lot of things that shipped and so specifically what we want to look at is the multimodality of Gemini and the gen media models Uh so actually this slide is already out to date because this morning we shipped a new model uh that is called anyone know which model did we ship? We shipped Omni Flash which is uh a video generation and editing model and we shipped Nano Banana Light. So Nano Banana Light is a really exciting model because it cuts down the latency of image generation and image editing quite significantly. And so what I've done, we'll we'll do a couple demos here. Uh and one of the demos is um this example here, which I call the uh Gemini Nanobanana story canvas. Uh and I just VIP code this up uh just recently and I'm having some trouble with my mouse here. Um but so what we can do is we can come up with a story with um Gemini Life and then have it use Nano Banana Light to to illustrate that story for us. So anyone you know this is a little bit like improv theater now. Anyone has a story line they want to start with? Where are we somewhere? Okay, let's say um we're in Taipei in the 1980s. The neon lights of Taipe's night market pulsed with an almost magical energy reflecting off the wet pavement and the chrome of passing scooters. Rain had just passed, leaving a humid mist that hung in the air. Amidst the throng of shoppers and vendors, a young woman clutched a small, strangely glowing object under her jacket. Something very exciting happened in 1980s Taipei. A new kind of drink emerged. She paused at a stall where a man was vigorously shaking a metal cup, creating a frothy, milky concoction. He handed a cup to a customer who took a long sip with wide eyes, savoring the strange tapioca pearls at the bottom. This was the birth of bubble tea, and its sweet, chewy magic was spreading like wildfire. So what you can see here is uh from the tool call. So Gemini Life is getting a tool call um and then it's using the Gemini interactions API to generate an image on the fly. Now it still takes kind of you know two to three seconds for the actual image generation but that is um pretty impressively fast. And um what you can look at is if you're looking at the code here uh we're using the interactions API. And so the interactions API is exciting because it is a stateful API which means we can chain our interactions very seamlessly. So you could see here the first image generation that we were doing uh it didn't have a previous interaction ID but then the next generation that we were doing we're passing this previous interaction ID to the interactions API which means that we have the previous context of the previous interaction and we can continue that in the generation. So what that means specifically for our demo here is that you can see is our first you know interaction uh it generated an image with a very specific kind of visual feeling and also with characters right and so we can continue that interaction and so we can keep the same consistency kind of across these generations. So if you haven't tried out the interactions API, uh it is our unified API for interacting with all of the Google DeepMind models as well as agents. Uh and so do try that out. It is a it is a really really cool um way to interact with everything that Google DeepMind has. Okay, so this was you know a bunch of things that happened. Um so this was kind of on the Gen Media side. We shipped uh Nano Banana Light this morning. Now, we also recently shipped on the audio side our live translation model. Has anyone tried out Gemini 3.5 live translate? Yes. Okay, a couple folks. Uh, that's fantastic. So, what we can do is uh we can try that out as well. We can play uh a little demo. And so what I'm going to do here is I'm going to create this live translation session and I want you to scan uh this QR code on your phone. And so what this is doing here is if I now enable um my microphone. So this is getting the the microphone of my speech um here into the session. And now if you go in and you select your preferred language. So we're seeing here we're getting some German. So someone is listening to me now speaking in German. And so if you are uh now listening to this, so you probably need some headphones. So if you have some noiseancelling headphones, you can now listen into this talk in your preferred language. And so Gemini live translate is a really incredible model. It's near realtime translation. So as I'm speaking, as soon as the model is recognizing kind of the context of what I'm speaking about, uh it will then start translating in real time to your desired language. So I see someone is listening in there. Do do you get some audio out? Can you hear it? Yeah. Does it sound good? Fantastic. See, um, so this is this is really exciting because you know in a conference like this obviously we have a very very international audience and so I can speak in my preferred language. So I could switch to German now for example. the powerful Gemini live translates. Um, it is a a really incredible model. Um, it automatically detects the language the speaker is speaking and then translates it in real time uh into our preferred language. Okay, we have uh a very international crowd here today. Uh, who's listening in in uh Chinese? Are you listening in Chinese? Oh, okay. I didn't understand that, but okay. Um, so that is Gemini live translate. Um, if you want to try it out, you can add it to your um applications now. It is available on the Gemini API. So, do try that out. uh it is a really incredible model and uh would love to hear kind of what you're what you're building with that. Now obviously we're looking at uh agents, right? Can your agents uh what's it the title of this talk? Gemini made up this title. So I uh can your agent hear me now? I think that that is the title. So our most recent real time conversational model is called Gemini 3.1 Flash Life. Now, if you've talked to Gemini before in the Gemini app, that is exactly the model that you're speaking to. And so, we can do a little demo here. So, this demo was vipcoded by Gemini in Google AI Studio. And if you remember the good old times, you know, when you were listening to radio, you know, not too long ago, you could phone up the the the radio host and you could ask for a song, right? You could wish for a song. So, it's the same idea here. We're using Gemini Life to be our radio DJ and then we're giving Gemini Life an API, a function call to the Gemini API to then use LIA 3, which is our music generation model to then generate some music for us on demand. Okay, let's uh try this out. We're we're just going live demo to live demo here. Let's let's see. Well, hello there. You're live on the live jukebox right now. What absolute banger are we cooking up for you today, mate? To get this track properly sorted, I need a bit more of an idea. What genre are we thinking? Upbeat synth pop, smooth jazz, something with a bit more well. And what's the vibe lyrics-wise? Give us some details. \>\> Yeah. Can we get like a really high energy ragaton, please? About the AI engineering world's fair here in San Francisco. \>\> Now that sounds like a proper belter, mate. high energy regaton about the AI engineering world's fair. Brilliant. Right. To make this absolutely top of the pops, any specific bits of tech or big names you want thrown in, or are we just going for that huge festival vibe? \>\> Yeah, just surprise me, please. You got it, my friend. Prepare for a huge musical surprise generating that track right now. Mixing that regaton rhythm with the tech buzz for the World's Fair. It's going to blow your headphones off. Fair. What a feeling. Let's go on the floor. Neural networks firing up. Groundbreaking for sure. Clever algorithms dancing. Tech pioneers mingling. This is innovation. Baseline is pulsing heavy. Can you feel the heat? Sin stabs electrifying. Move in your feet. This is how we are. Yeah, that that was pretty cool. You went multilingual there as well. Nice one. Cheers. \>\> Cheers, mate. Glad you like the flavor. Keep those big ideas coming. This jukebox never sleeps. What are we spinning next then? Feeling. \>\> All right. Um, yes, you can try this out yourself. Uh, it is is a a demo that was built in Google AI Studio. Uh so if you want you can fork this in AI studio and continue yourself. Um I know this application might not necessarily be that useful for your business context, you know, if that's something you're looking to do, but I think the individual components are definitely useful. So if you're building real-time conversational agents, uh you know, you want them to be multilingual, you want them to have capabilities, uh tool calls. So in this example, we were just giving the agent a tool call to generate music with LIA 3 uh on demand. Now obviously this could also just call out to your CRM and get some you know customer information or it could create you know a new ticket on the back end. Um so you know really what I want you to take away is kind of this real time multimodal conversational uh experience that you can build. And so the way that you can build with this model, so the the Gemini 3.1 flash real-time conversational model is through the life API. So the life API is a low latency real time voice and vision uh API. So you can build multiodal uh life real-time interactions. And so what this means is you have this stateful websocket API where you create a session and then over this websocket you can send in real time text, audio and video frames to the model and the model will then respond in real time back in audio chunks. And these audio chunks you can then stream uh in your application. Uh and in addition to the audio, you also get the transcript of what was said. So if you've worked with voice agents before, this is a little bit different from kind of your cascading voice agent pipeline because this is a native audio model. So the model itself does not go through text, but rather the intelligence is baked into the audio model. So the audio model can understand the speech naturally and natively and can then respond back in any of the 90 supported languages. So this is what allows us to to switch very fluidly between different languages. Um so kind of the key features here, yeah, multilingual support. So 90 uh different languages currently supported in the preview. Um we can barge in, we can interrupt the model, right? If our DJ is talking a bit too lengthy, uh we we can we can barge in, interrupt it, we can give it uh the tools and we have the audio transcription as well. Now, if you're building real time conversational applications or real-time um multimodal applications, uh you know, real-time media can be quite challenging to scale. So, we have a great lineup of partners. So, you know, LifeKit, Pipecat, they all actually here in the expo as well. I recommend you go visit them. They do some really really cool stuff to allow you to get up and running quickly with the Gemini Life API. Okay, so if you're now um you know, you're like, "Okay, this sounds great. I want this." Um the easiest way to try this out is through Google AI Studio. So we spend a lot of money on this domain, AI.studio. So I do appreciate if you use that one. Um so we can get some ROI there. So ai.studioife. You can try out the model. You can talk to it and you can understand the capabilities that are available. So let's do that now. Uh so we're just going to AI.dev. We also have AI.dev. Um yeah, please do use that as well. That one was was also not not so cheap. Uh there we go. AI.devlife. And so here uh we can also try out the live translate model if you want to. Uh but we can here select our Gemini 3.1 flash uh live model. We can select a different voice. Uh for example, I like puck here. And then also we can enable grounding with Google search. So obviously the model is trained on a lot of information but it does not have access to real time information, right? And so by giving it access to Google search, it can actually go off and it's basically like a built-in function call. And if we ask it about today's weather or you know why Germany played so well in the World Cup, uh we can we can do that and it can use Google to to look that up. Okay, let's give this a try. \>\> Hey, can you hear me? \>\> Yes, I can hear you perfectly. How can I help you today? \>\> Can you see me? Yes, I can see you clearly. You're wearing a black shirt and a black hat sitting in front of a blue background. Anything specific you wanted to show me? \>\> Yeah. What does it say on my sweatshirt? \>\> I can see the word Gemini printed in yellow letters. Anything else you'd like me to identify? Uh, okay. I think my uh my Mandarin didn't work out so well there. Uh, can you also speak Mandarin, please? \>\> Yes, I can speak Mandarin, too. What would you like to say now? What's about? \>\> All right. If you don't read Mandarin, I do apologize. I was just asking the model what it has eaten, and I just said it doesn't need to eat because it's it's a model. So, uh, yeah, there we go. All right. So this is the easiest way for you to get started with uh Gemini ai.studiolife. If you want to learn more about building with Gemini and the Gemini live API, join me tomorrow morning in uh room 2009, I think. Check the schedule. It it does change every once in a while. Uh but other than that, Dankur, thank you. Cheers.