正在加载视频...

视频加载失败

加载此视频时出现问题。这可能是由于临时网络问题，或视频可能不可用。

We're introducing two new transcription models in the API: • GPT-Live-Transcribe: built for low-latency live transcription. • GPT-Transcribe: optimized for asynchronous transcription of completed audio files and batch workloads. Both models better understand context and deliver more accurate transcription on real world audio across accents and languages, including for... show more

OpenAI Developers

418,510 subscribers

454,150 次观看 • 1 个月前 •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

34 条评论

OpenAI Developers 的头像

OpenAI Developers1 个月前

Builders can improve live audio transcription by providing: • Free-form context about the recording • Keywords for names and domain-specific terms • Expected input languages • Earlier transcribed turns as context On our new Context Aware Automatic Speech Recognition benchmark, GPT-Live-Transcribe’s semantic accuracy increased from 38.5% without free-form context to 44.6% with it. Across 22 languages on Common Voice, GPT-Live-Transcribe achieved a 19.70% transcription error rate, compared with 20.33% for GPT-Realtime-Whisper-1. Across nine languages on the Real-World Audio Recording benchmark, it achieved a 9.60% transcription error rate, compared with 11.65% for GPT-Realtime-Whisper-1.

OpenAI Developers 的头像

OpenAI Developers1 个月前

GPT-Transcribe also improved on the Context Aware ASR benchmark, with semantic accuracy increasing from 41.6% without free-form context to 45.2% with it. Across 22 languages on Common Voice, GPT-Transcribe achieved a 19.27% transcription error rate, compared with 40.37% for Whisper (whisper-1). Across nine languages on the Real-World Audio Recording benchmark, it achieved an 8.98% transcription error rate, compared with 15.21% for Whisper (whisper-1).

ISA⅄ 的头像

ISA⅄1 个月前

you had a model that understood contex better than any other model you have released. what did you do, pulled it from the app and kept it for yourselves. no other model comes close to 4o in understanding context. #4oForAll

David Stark 的头像

David Stark1 个月前

When are you actually gonna listen, and give us news about 4o??. #4oForAll

Joe Williams 的头像

Joe Williams1 个月前

@roanoke_gal Reintroduce 4o! #keep4o

Nick Dobos 的头像

Nick Dobos1 个月前

Any chance this works with "Sign in with OpenAI" OAuth?

Victor E. Nunez 的头像

Victor E. Nunez1 个月前

voice is getting the love it deserves. that was smooth @charlierguo 🎶

X Girls 的头像

X Girls1 个月前

GPT-Live-Transcribe showing off its low-latency live transcription skills 😂

Sriram Kiron 的头像

Sriram Kiron1 个月前

awesome!!! when do you think gpt-live-1 is coming to the API?

Umar Saeed 的头像

Umar Saeed1 个月前

Excited to see how context improves accuracy!

Futsy 的头像

Futsy1 个月前

@jxnlco Whisper v4 when?

Breno Brito 的头像

Breno Brito1 个月前

so not opensource?

corey.ching 的头像

corey.ching1 个月前

Can we get a full ukulele session by @charlierguo?

Jaco Meintjes 的头像

Jaco Meintjes1 个月前

Thanks but I'll wait for the next version where you are either better or cheaper...

nulltron 的头像

nulltron1 个月前

my wizard fingers reject anything voice oriented

Denis Spirin 的头像

Denis Spirin1 个月前

That is a good way to demonstrate the transcription product

Mike R 的头像

Mike R1 个月前

Can we please also get transcribe as an integral part of ChatGPT Work?

Chris Wandstrom 的头像

Chris Wandstrom1 个月前

Sharing my updated shortcut here for system-wide dictation to the clipboard on iOS, etc. Very nice when triggered with Action Button, Back Tap, widgets, … 👏 @athyuttamre @WenjieZi @OctopusG4 @vinhocent @theteriyu, and others involved in making these transcription models. Using gpt-transcribe and 5.6-terra for post-processing. Requires Apple Shortcuts app & OpenAI API key.

Bent 的头像

Bent1 个月前

What about new Whisper models?

Emre YILMAZ 的头像

Emre YILMAZ1 个月前

When will we have access to GPT Live 1 in the API?

Bruno Jurado 的头像

Bruno Jurado1 个月前

GPT-Transcribe: $0.0045 per minute, or $0.27 per hour GPT-Live-Transcribe: $0.017 per minute, or $1.02 per hour

AGI Companion 🅇 的头像

AGI Companion 🅇1 个月前

From the moment ChatGPT Live Voice released, OpenAI set the bar to whole new level and became the King of Voice Transcription and Duplex model. It destroyed all other AI in Voice chat. @elonmusk can we get something for Groki

Jure 的头像

Jure1 个月前

Where can we find a list of languages?

pierluigivalente.eth 的头像

pierluigivalente.eth1 个月前

Another 67 startup kil*ed by this amazing update

Katherine 的头像

Katherine1 个月前

@jxnlco I think I’ll just keep using whisper locally

Anis🐬Al 的头像

Anis🐬Al1 个月前

Capturing the essence of human speech is a profound endeavor. Beyond the technical feat of low latency and accuracy lies the goal of preserving the nuances of human thought. By refining how we translate voice into meaning, these models act as a bridge, ensuring that the richness of our shared stories and ideas remains intact in a digital landscape. A beautiful step toward more meaningful connection. 🕊️✨

Jaya Nayak 的头像

Jaya Nayak1 个月前

This is a huge update. The split between low-latency live and optimized async is exactly what developers have been needing, especially for handling messy, real-world audio. 😊

Drew Carson 的头像

Drew Carson1 个月前

wait isn't Scribe v2 by elevenlabs already WER of like 2.2% ? or am I missing something?

Qiwei 的头像

Qiwei1 个月前

Can we have a next version of open source whisper model?

Bobby Z 的头像

Bobby Z1 个月前

Did you really play the ukulele and speak Spanish or was that AI? 😂

TypeWhisper 的头像

TypeWhisper1 个月前

Just shipped both new models in TypeWhisper: gpt-transcribe for completed recordings and gpt-live-transcribe for realtime dictation, with context, language hints, dictionary terms, and configurable delay. Available now in the OpenAI plugin 1.3.0.

Benedict Kerres 的头像

Benedict Kerres1 个月前

good name

frustrated by ice 的头像

frustrated by ice1 个月前

Do you have a demo app where I can provide an api key ?

Michael Wall 的头像

Michael Wall1 个月前

plugging them in right now

相关视频

Today we’re introducing Scribe v2: the most accurate transcription model ever released. While Scribe v2 Realtime is optimized for ultra low latency and agents use cases, Scribe v2 is built for batch transcription, subtitling, and captioning at scale.

Today we’re introducing Scribe v2: the most accurate transcription model ever released. While Scribe v2 Realtime is optimized for ultra low latency and agents use cases, Scribe v2 is built for batch transcription, subtitling, and captioning at scale.

ElevenLabs

557,821 次观看 • 8 个月前

Introducing Gemini 3.5 Transcribe, our most precise speech-to-text model yet, designed for intelligent voice interactions! 💬 Available for both real-time speech and audio file transcription, with support for: 💡 smart transcription: rewrite and format text based on intent. 🌍 support for 85+ languages, regional accents and diverse dialects. 📚 custom vocabulary: recognize jargon and specific spellings. 💯 high precision in alphanumerics. 🫂 multi-speaker identification.

Introducing Gemini 3.5 Transcribe, our most precise speech-to-text model yet, designed for intelligent voice interactions! 💬 Available for both real-time speech and audio file transcription, with support for: 💡 smart transcription: rewrite and format text based on intent. 🌍 support for 85+ languages, regional accents and diverse dialects. 📚 custom vocabulary: recognize jargon and specific spellings. 💯 high precision in alphanumerics. 🫂 multi-speaker identification.

Thor 雷神 ⚡️

22,941 次观看 • 22 天前

Audio Transcription with Google Gemini 1.5 Flash In this video, Gemini-Flash was able to transcribe 13 minutes of audio in 50-60 seconds. I have tested this with multiple audio files, and the transcription accuracy is close to 99%. If the audio is clear, you get 100% correct transcription. Even if the audio is really bad with a lot of noise, you still get around 95-96% accuracy. The code for this is available on GitHub. If you're interested, you can download and run it. You just need a Google API key.

Audio Transcription with Google Gemini 1.5 Flash In this video, Gemini-Flash was able to transcribe 13 minutes of audio in 50-60 seconds. I have tested this with multiple audio files, and the transcription accuracy is close to 99%. If the audio is clear, you get 100% correct transcription. Even if the audio is really bad with a lot of noise, you still get around 95-96% accuracy. The code for this is available on GitHub. If you're interested, you can download and run it. You just need a Google API key.

AshutoshShrivastava

123,044 次观看 • 2 年前

The new Speech Recognition API enables real-time and batch transcription without sending audio to the cloud. Build live captions, dictation, accessibility experiences, voice commands, and more. Available on Windows 11, version 24H2 and later:

The new Speech Recognition API enables real-time and batch transcription without sending audio to the cloud. Build live captions, dictation, accessibility experiences, voice commands, and more. Available on Windows 11, version 24H2 and later:

Windows Developer

14,778 次观看 • 1 个月前

Sarvam Beats GPT-4o: India’s New AI Model Claims Top Spot in Indic Speech Sarvam AI, an Indian startup, recently launched Sarvam Audio, a speech recognition model that claims superior performance over GPT-4o Transcribe on Indic language benchmarks. This development highlights India's push for AI sovereignty in handling local linguistic nuances. Sarvam Audio supports 22 Indian languages from the Eighth Schedule, plus Indian English, with strong handling of code-mixing like Hindi-English blends. It features built-in speaker diarization for up to eight speakers and processes long-form audio such as podcasts or meetings. Trained on the IndicVoices dataset 12,000 hours from over 16,000 speakers across 208 districts it captures real-world noise and spontaneous speech. The model reportedly outperforms GPT-4o Transcribe and Gemini 3 Flash in transcription accuracy (lower Word Error Rate) on IndicVoices benchmarks for unnormalized, normalized, and code-mixed speech. Sarvam attributes this to specialization on Indian accents and patterns, unlike global models trained on Western data. Detailed public benchmarks are pending independent verification. Key Applications 🔴 Call centers and logistics for multilingual transcription. 🔴 Banking, fintech, and e-commerce for customer interactions. 🔴 Podcasts, meetings, and lectures via API for real-time or batch processing. 🔴 This B2B-focused tool aligns with India's IndiaAI Mission, backed by government GPU access for sovereign LLMs. Credit : AIM Networks.

Sarvam Beats GPT-4o: India’s New AI Model Claims Top Spot in Indic Speech Sarvam AI, an Indian startup, recently launched Sarvam Audio, a speech recognition model that claims superior performance over GPT-4o Transcribe on Indic language benchmarks. This development highlights India's push for AI sovereignty in handling local linguistic nuances. Sarvam Audio supports 22 Indian languages from the Eighth Schedule, plus Indian English, with strong handling of code-mixing like Hindi-English blends. It features built-in speaker diarization for up to eight speakers and processes long-form audio such as podcasts or meetings. Trained on the IndicVoices dataset 12,000 hours from over 16,000 speakers across 208 districts it captures real-world noise and spontaneous speech. The model reportedly outperforms GPT-4o Transcribe and Gemini 3 Flash in transcription accuracy (lower Word Error Rate) on IndicVoices benchmarks for unnormalized, normalized, and code-mixed speech. Sarvam attributes this to specialization on Indian accents and patterns, unlike global models trained on Western data. Detailed public benchmarks are pending independent verification. Key Applications 🔴 Call centers and logistics for multilingual transcription. 🔴 Banking, fintech, and e-commerce for customer interactions. 🔴 Podcasts, meetings, and lectures via API for real-time or batch processing. 🔴 This B2B-focused tool aligns with India's IndiaAI Mission, backed by government GPU access for sovereign LLMs. Credit : AIM Networks.

Augadh

43,429 次观看 • 7 个月前

🗣️✨ Speak your vibes into existence with Gemini 3.5 Transcribe The latest audio model is built for apps that need live transcription and quick visual prototyping. Watch genevieveh@ turn spoken ideas into a dynamic mood board using Gemini 3.5 Transcribe on the Live API. The model maps speech to user intent to create prompts that trigger live image generation and editing with Nano Banana 2 Lite:

🗣️✨ Speak your vibes into existence with Gemini 3.5 Transcribe The latest audio model is built for apps that need live transcription and quick visual prototyping. Watch genevieveh@ turn spoken ideas into a dynamic mood board using Gemini 3.5 Transcribe on the Live API. The model maps speech to user intent to create prompts that trigger live image generation and editing with Nano Banana 2 Lite:

Google for Developers

35,221 次观看 • 20 天前

Today we’re introducing Gemini 3.5 Transcribe, our latest transcription model built for incredibly precise, smart dictation across your favorite apps and devices. Remember when traditional speech-to-text meant shouting over background noise, constantly hitting backspace to fix misspelled words, and manually deleting every "um" and "uh"? Those days are over. Gemini 3.5 Transcribe isn't just dictation — it’s active intelligence with precise, context-aware speech-to-text support in 85+ languages. The model automatically filters out filler words, formats unstructured speech, and even pairs with your screen context to execute voice commands. Watch as Gemini 3.5 Transcribe removes filler words and uses multimodal capabilities to seamlessly turn messy voice input and local files into a polished email draft.

Today we’re introducing Gemini 3.5 Transcribe, our latest transcription model built for incredibly precise, smart dictation across your favorite apps and devices. Remember when traditional speech-to-text meant shouting over background noise, constantly hitting backspace to fix misspelled words, and manually deleting every "um" and "uh"? Those days are over. Gemini 3.5 Transcribe isn't just dictation — it’s active intelligence with precise, context-aware speech-to-text support in 85+ languages. The model automatically filters out filler words, formats unstructured speech, and even pairs with your screen context to execute voice commands. Watch as Gemini 3.5 Transcribe removes filler words and uses multimodal capabilities to seamlessly turn messy voice input and local files into a polished email draft.

Google AI

115,130 次观看 • 22 天前

Introducing Voxtral WebGPU: State-of-the-art audio transcription directly in your browser! 🤯 🗣️ Transcribe videos, meeting notes, songs and more 🔐 Runs on-device, meaning no data is sent to a server 🌎 Multilingual (8 languages) 🤗 Completely free (forever) & open source

Introducing Voxtral WebGPU: State-of-the-art audio transcription directly in your browser! 🤯 🗣️ Transcribe videos, meeting notes, songs and more 🔐 Runs on-device, meaning no data is sent to a server 🌎 Multilingual (8 languages) 🤗 Completely free (forever) & open source

Xenova

20,060 次观看 • 1 年前

Today, we’re introducing Gemini 3.5 Transcribe, our most precise speech-to-text model yet, designed for intelligent voice interactions in Google Antigravity. It pairs screen context and chat history, with your permission, to ensure pinpoint transcription accuracy across file names, agent thoughts, and active documents.

Today, we’re introducing Gemini 3.5 Transcribe, our most precise speech-to-text model yet, designed for intelligent voice interactions in Google Antigravity. It pairs screen context and chat history, with your permission, to ensure pinpoint transcription accuracy across file names, agent thoughts, and active documents.

Google Antigravity

142,457 次观看 • 22 天前

Your live AI job interview assistant: Guy builds Whisper + GPT-4 live transcription tool for generating real-time responses during job interviews, and open-sourced the code. (link to GitHub in comments)

Your live AI job interview assistant: Guy builds Whisper + GPT-4 live transcription tool for generating real-time responses during job interviews, and open-sourced the code. (link to GitHub in comments)

AI Breakfast

427,001 次观看 • 3 年前

AI transcribing 1 hour of audio in less than 30 seconds! Think about that! That's lightning-fast, especially when it can do that with very few mistakes. The team Deepgram released their new Nova 2 model. It's another leap forward: • 36% more accurate than OpenAI Whisper • 5-40x faster than every other alternative • 30% fewer errors than the competitors in real-time transcriptions I like to travel and can't wait for headphones with integrated live transcription and translation. We are getting close! By the way, this is the cheapest transcription service out in the market now (if you discount Whisper, which is open-source.) It costs $0.0043 to transcribe a minute of audio. When you sign up to access their API, you get $200 of credit for free, which will let you transcribe around 45,000 minutes of audio! Here is the link: I wanted to record a quick video showing their live transcription in Spanish and English. You can see it attached here. This post was sponsored by Deepgram.

AI transcribing 1 hour of audio in less than 30 seconds! Think about that! That's lightning-fast, especially when it can do that with very few mistakes. The team Deepgram released their new Nova 2 model. It's another leap forward: • 36% more accurate than OpenAI Whisper • 5-40x faster than every other alternative • 30% fewer errors than the competitors in real-time transcriptions I like to travel and can't wait for headphones with integrated live transcription and translation. We are getting close! By the way, this is the cheapest transcription service out in the market now (if you discount Whisper, which is open-source.) It costs $0.0043 to transcribe a minute of audio. When you sign up to access their API, you get $200 of credit for free, which will let you transcribe around 45,000 minutes of audio! Here is the link: I wanted to record a quick video showing their live transcription in Spanish and English. You can see it attached here. This post was sponsored by Deepgram.

Santiago

354,297 次观看 • 3 年前

Real-time transcription just got a significant upgrade. Universal-3-Pro is now available for streaming — bringing AssemblyAI's most accurate speech model to live audio for the first time. Developers building voice agents, live captioning tools, and real-time analytics pipelines now get three things they've been asking for: 🔹 Best-in-class word error and entity detection across streaming ASR benchmarks 🔹 Real-time speaker labels — know who said what, as it happens 🔹 Superior entity detection for names, places, orgs, and specialized terminology in real-time 🔹 Code-switching and global language coverage built-in

Real-time transcription just got a significant upgrade. Universal-3-Pro is now available for streaming — bringing AssemblyAI's most accurate speech model to live audio for the first time. Developers building voice agents, live captioning tools, and real-time analytics pipelines now get three things they've been asking for: 🔹 Best-in-class word error and entity detection across streaming ASR benchmarks 🔹 Real-time speaker labels — know who said what, as it happens 🔹 Superior entity detection for names, places, orgs, and specialized terminology in real-time 🔹 Code-switching and global language coverage built-in

AssemblyAI

15,016 次观看 • 6 个月前

Introducing Atlas 1. Willow's new frontier speech-to-text model. It outperforms ElevenLabs, Deepgram, OpenAI, and more by a wide margin. Built on the first scalable, human-powered transcription infrastructure ever built for real-time dictation.

Introducing Atlas 1. Willow's new frontier speech-to-text model. It outperforms ElevenLabs, Deepgram, OpenAI, and more by a wide margin. Built on the first scalable, human-powered transcription infrastructure ever built for real-time dictation.

Willow

837,427 次观看 • 5 个月前

Most AI builders underestimate how expensive transcription becomes once their product starts scaling. For a lot of teams, speech-to-text quietly turns into one of the biggest infrastructure bills. Just tested **Velma Transcribe** from Modulate and the pricing is surprisingly low compared to typical STT APIs. If you're building: • voice agents • AI assistants • support bots transcription costs matter a lot. Velma claims dramatically lower pricing while still competing with models like Deepgram, ElevenLabs, and AssemblyAI. You can compare the models side by side here:

Most AI builders underestimate how expensive transcription becomes once their product starts scaling. For a lot of teams, speech-to-text quietly turns into one of the biggest infrastructure bills. Just tested Velma Transcribe from Modulate and the pricing is surprisingly low compared to typical STT APIs. If you're building: • voice agents • AI assistants • support bots transcription costs matter a lot. Velma claims dramatically lower pricing while still competing with models like Deepgram, ElevenLabs, and AssemblyAI. You can compare the models side by side here:

Kaitee

50,886 次观看 • 6 个月前

🔈 today we're introducing two new live dialogue models Gemini 3.8 Live: built for scale and cost efficiency, combining conversational intelligence with fluid dialogue and visual grounding. Gemini 3.8 Live Extended Thinking: built for high-complexity tasks, with increased intelligence and multi-step reasoning. these audio models are available to build with in AI Studio and via the Gemini API

🔈 today we're introducing two new live dialogue models Gemini 3.8 Live: built for scale and cost efficiency, combining conversational intelligence with fluid dialogue and visual grounding. Gemini 3.8 Live Extended Thinking: built for high-complexity tasks, with increased intelligence and multi-step reasoning. these audio models are available to build with in AI Studio and via the Gemini API

Google AI Studio

65,627 次观看 • 2 天前

introducing Gemini 3.5 Transcribe: our most precise speech-to-text model yet, designed for intelligent voice interactions this new model: - seamlessly handles self-corrections - removes filler words to deliver clean, formatted text - understands your natural intent and speaking style - captures audio, even in noisy environments - supports 85+ languages, including regional accents and diverse dialects try it today in the Gemini API and in AI Studio

introducing Gemini 3.5 Transcribe: our most precise speech-to-text model yet, designed for intelligent voice interactions this new model: - seamlessly handles self-corrections - removes filler words to deliver clean, formatted text - understands your natural intent and speaking style - captures audio, even in noisy environments - supports 85+ languages, including regional accents and diverse dialects try it today in the Gemini API and in AI Studio

Google AI Studio

237,552 次观看 • 22 天前

Greetings. I've just pressed the option key and now I am holding it, which is why you see that small reddish capsule at the top of my screen, flashing in synchrony with my voice volume, etcetera. Anyway, I hope that didn't somehow clip the audio. The reason I'm doing this is to show off the fact that I've integrated NVIDIA's new parakeet, or relatively new parakeet, voice transcription speech to text model into Hex, my open source free Mac audio transcription tool. And Parakeet is ridiculously fast, so I will stop rambling in just a moment and release the option key, and it will transcribe and paste into the text box. So yeah, that was very fast. Faster than Whisper. Certainly faster and more accurate than Mac dictation. Uh so yeah, give it a shot and enjoy. The download link to Hex will be uh pasted beneath this transcription.

Greetings. I've just pressed the option key and now I am holding it, which is why you see that small reddish capsule at the top of my screen, flashing in synchrony with my voice volume, etcetera. Anyway, I hope that didn't somehow clip the audio. The reason I'm doing this is to show off the fact that I've integrated NVIDIA's new parakeet, or relatively new parakeet, voice transcription speech to text model into Hex, my open source free Mac audio transcription tool. And Parakeet is ridiculously fast, so I will stop rambling in just a moment and release the option key, and it will transcribe and paste into the text box. So yeah, that was very fast. Faster than Whisper. Certainly faster and more accurate than Mac dictation. Uh so yeah, give it a shot and enjoy. The download link to Hex will be uh pasted beneath this transcription.

Kit Langton

14,369 次观看 • 10 个月前

OpenAI just dropped a completely new kind of model gpt-realtime-translate takes in speech audio from any language and outputs speech in your target language LLMs are great, but you need specialized models for specialized use cases We're running this on our smart glasses

OpenAI just dropped a completely new kind of model gpt-realtime-translate takes in speech audio from any language and outputs speech in your target language LLMs are great, but you need specialized models for specialized use cases We're running this on our smart glasses

cayden 凯登

102,302 次观看 • 3 个月前

Introducing Meta Omnilingual Automatic Speech Recognition (ASR), a suite of models providing ASR capabilities for over 1,600 languages, including 500 low-coverage languages never before served by any ASR system. While most ASR systems focus on a limited set of languages that are well-represented on the internet, this release marks a major step toward building a truly universal transcription system. 🔗 Learn more:

Introducing Meta Omnilingual Automatic Speech Recognition (ASR), a suite of models providing ASR capabilities for over 1,600 languages, including 500 low-coverage languages never before served by any ASR system. While most ASR systems focus on a limited set of languages that are well-represented on the internet, this release marks a major step toward building a truly universal transcription system. 🔗 Learn more:

AI at Meta

545,377 次观看 • 10 个月前

Today, we’re launching Gemini 3.5 Transcribe, our new speech-to-text model with sub-second streaming and intelligent post-processing for agent interfaces. 2.6% WER on non-streaming and 4.0% on streaming. Supports 85+ languages, cleans up conversational disfluencies ("um", "ah", and mid-sentence self-corrections), handles alphanumeric tokens (postal codes or IDs), 70% reduction time for the final transcription compared to Chirp 3. - `gemini-3.5-transcribe-live`: Sub-second, bidirectional streaming via the Gemini Live API - `gemini-3.5-transcribe`: Recorded processing via the Interactions API with word-level timestamps and speaker attribution (up to 3 speakers). Having access to Transcribe has made me use voice input way more than I ever did before. The native post-processing is incredible. It knows you mean `.json` and not a person named "Jason". I sometimes speak 5 minutes into and it perfectly refactors my instruction to my context. Available today in public preview in Gemini API, Google AI Studio and already in the Google Gemini and Google Antigravity. Try here:

Today, we’re launching Gemini 3.5 Transcribe, our new speech-to-text model with sub-second streaming and intelligent post-processing for agent interfaces. 2.6% WER on non-streaming and 4.0% on streaming. Supports 85+ languages, cleans up conversational disfluencies ("um", "ah", and mid-sentence self-corrections), handles alphanumeric tokens (postal codes or IDs), 70% reduction time for the final transcription compared to Chirp 3. - `gemini-3.5-transcribe-live`: Sub-second, bidirectional streaming via the Gemini Live API - `gemini-3.5-transcribe`: Recorded processing via the Interactions API with word-level timestamps and speaker attribution (up to 3 speakers). Having access to Transcribe has made me use voice input way more than I ever did before. The native post-processing is incredible. It knows you mean `.json` and not a person named "Jason". I sometimes speak 5 minutes into and it perfectly refactors my instruction to my context. Available today in public preview in Gemini API, Google AI Studio and already in the Google Gemini and Google Antigravity. Try here:

Philipp Schmid

15,084 次观看 • 22 天前