Superpower Daily
EMXETF Lists TGRZ, a China LLM ETF With an 80% China-Exposure M...Anthropic Opens 750,000 Claude Conversations to Outside Study W...AI Pact Draws 15-Plus Candidates as Data-Center Backlash Reache...Amazon’s Las Vegas Warehouse Reportedly Cuts Up Books for AI Tr...Harvard Uses 7 AI Faculty Avatars in a $699 Founder Bootcamp, W...Hugging Face Explains DeepSeek’s Matched Shift From MoE Experts...EMXETF Lists TGRZ, a China LLM ETF With an 80% China-Exposure M...Anthropic Opens 750,000 Claude Conversations to Outside Study W...AI Pact Draws 15-Plus Candidates as Data-Center Backlash Reache...Amazon’s Las Vegas Warehouse Reportedly Cuts Up Books for AI Tr...Harvard Uses 7 AI Faculty Avatars in a $699 Founder Bootcamp, W...Hugging Face Explains DeepSeek’s Matched Shift From MoE Experts...
Modelspublished

Google Puts Gemini 3.5 Transcribe in Preview With Live and Recorded-Audio APIs

The release turns speech recognition into a Gemini product layer for developers and Google surfaces, but a product-specific price remains unavailable for teams weighing production use.

By 3 min read
Google Puts Gemini 3.5 Transcribe in Preview With Live and Recorded-Audio APIs
Sign in to save and react

Listen to this story

The audio brief

About 1:28
0:001:28
Read transcript
Google has put Gemini 3.5 Transcribe into public preview, giving developers two separate ways to turn audio into text: one for live interaction, and one for completed recordings. The distinction is practical. The Live API uses gemini-3.5-transcribe-live as a continuous, two-way audio stream, with Google claiming sub-second latency for voice agents and live captions. For meetings, call logs, and other recordings, the Interactions API uses gemini-3.5-transcribe. That mode adds word-level timestamps and speaker attribution for up to three people. Support beyond three speakers is still experimental, which limits the preview for larger meetings and group calls. The system can also remove filler words, resolve spoken corrections, format the result, and use a custom vocabulary. Google says it handles alphanumeric details such as postal codes and order IDs, across more than 85 languages, including regional accents and dialects. Access is available through Google AI Studio, Google Antigravity, and the Gemini Enterprise Agent Platform preview. Google reports word error rates of 4.0 percent for streaming and 2.6 percent for non-streaming, plus a 70 percent improvement in final-transcription latency over Chirp 3. Those are launch figures, and the biggest production question remains unanswered: Google has not published a product-specific price for Gemini 3.5 Transcribe.

Story brief

3 key points

Google’s public preview of Gemini 3.5 Transcribe gives builders a choice between low-latency streaming for live voice products and richer processing for completed recordings. The latter can attach word-level timestamps and identify up to three speakers, while the system can clean filler words and handle alphanumeric terms across more than 85 languages. Google reports 4.0% streaming and 2.6% non-streaming word error...

  1. 01

    The Live API uses gemini-3.5-transcribe-live for continuous bidirectional audio and Google-stated sub-second latency.

  2. 02

    The Interactions API uses gemini-3.5-transcribe for recordings, with timestamps and speaker attribution for up to three speakers.

  3. 03

    Speaker support beyond three people is experimental, limiting immediate use for larger meetings or group calls.

Google has released Gemini 3.5 Transcribe in public preview, giving developers separate APIs for real-time voice interaction and recorded-audio processing. The split matters because a live assistant needs speed, while a call archive needs speaker labels and timestamps that can be searched or routed downstream.

One model, two operating modes

For interactive applications, the Live API exposes gemini-3.5-transcribe-live as a continuous, bidirectional stream with Google-stated sub-second latency. That is the route for voice agents and live captioning, where the system must keep accepting audio as it returns results.

The Interactions API uses a different model name, gemini-3.5-transcribe, for meetings, call logs, and other completed recordings. Its added outputs are word-level timestamps and attribution for as many as three speakers; Google labels support beyond three speakers experimental.

The API choice follows the job

01gemini-3.5-transcribe-live

Live API

Google positions the Live API for continuous bidirectional transcription in interactive voice applications, with stated sub-second latency.

02gemini-3.5-transcribe

Interactions API

Google positions the Interactions API for completed audio and includes speaker attribution and word-level timestamps.

Recognition is only part of the output

The model is designed to do more than render utterances word for word. Google says it can remove filler words, resolve a spoken self-correction, format the result, and adapt to a supplied custom vocabulary. That makes the output a cleaned-up working draft, not necessarily a verbatim record.

Google also emphasizes terms that frequently break automated workflows: postal codes, order IDs, and other alphanumeric entities, including in noisy recordings. It says the model automatically detects and transcribes more than 85 languages, and can handle regional accents and dialects.

Google is distributing the capability across several surfaces

  • Developers can access the public preview through the Gemini API in Google AI Studio and Google Antigravity, while enterprises can use the Gemini Enterprise Agent Platform preview.
  • Consumer availability includes the Gemini app on macOS in English and Rambler on Android in selected countries and languages; Chrome support is planned.
  • On Android, Rambler uses the transcription system for formatted dictation and voice-led edits such as correcting misspellings or changing writing style.

Performance claims meet an incomplete buying picture

Google says Artificial Analysis measured average word error rates of 4.0% for streaming and 2.6% for non-streaming use. Google also says time to final transcription improved 70% over Chirp 3, its prior transcription model. Those results distinguish the two modes on accuracy, but they remain performance figures presented in Google’s launch material.

The launch does establish model identifiers and API paths, addressing the practical question of how each mode is called. But Google’s pricing page does not list a price for Gemini 3.5 Transcribe. For a preview aimed at voice agents, captioning tools, and post-call analytics, that leaves the cost comparison unresolved.

Editorial analysis

Our Read

Google’s pitch is not simply that it can recognize speech. It is that transcription can become an intent-aware interface: clean up dictation, preserve specialized terms, and feed voice-driven actions across its products. The more consequential test is commercial clarity. Google has supplied named API models and preview routes, but teams building high-volume call, captioning, or agent systems will need a product-specific price and operating terms before they can compare this service against other transcription options. Those details are the next meaningful signal to watch.

Sources

  1. blog.googleIntelligent transcription with Gemini 3.5 Transcribe

Related coverage

Continue reading