Google has revealed Gemini 3.5 Transcribe, a new AI audio model that it says is built for higher-precision speech recognition and transcription, according to Engadget. The model joins Gemini 3.5 Live and Gemini 3.5 Live Experimental in the Gemini Audio family. The pitch is not just raw transcription. Engadget reports that Google says Gemini 3.5 Transcribe can adapt unstructured speech into cleaner formatted text, including handling self-corrections and removing filler words. The model is also described as supporting voice-based edits, so users can dictate and revise without manually typing each change. Google says the model can automatically detect more than 85 languages, according to the report. It also claims Gemini 3.5 Transcribe can learn custom vocabulary and unique spellings, a feature aimed at the proper nouns, internal terms and specialized words that often break generic speech-to-text systems. The model is also being positioned for more structured audio workflows. Engadget reports that Google says Gemini 3.5 Transcribe can capture alphanumeric strings such as order numbers and postal codes, and can attribute speech to as many as three speakers with word-level timestamps from pre-recorded audio. That would make the model relevant not only to dictation, but also to uses such as meeting, interview and podcast transcription, though the report does not provide performance benchmarks. The rollout spans consumer apps, developer tools and agentic systems. Engadget reports that Gemini 3.5 Transcribe powers the Rambler feature on Android devices including Pixel 11-series phones, as well as the Gemini app on macOS, where it can work with other Gemini models on agentic tasks. The model is also available in Google Antigravity, Google's agentic development platform. Chrome integration is still framed as forthcoming. Engadget's summary says Gemini 3.5 Transcribe will soon let users use speech-to-text in any web field in Chrome. The article body later uses the phrase 'text-to-speech,' but the examples it gives — dictating replies and posts and using voice to prompt Gemini — point to voice dictation in web fields rather than read-aloud output. Google also says the model is coming to Search Live, Gemini Live, Docs, Keep and Gmail, according to Engadget. Developers will be able to access Gemini 3.5 Transcribe through application programming interfaces, giving Google a path to push the same speech layer into third-party products as well as its own apps. Who benefits: Google benefits if Gemini 3.5 Transcribe makes voice input more useful across Chrome, Workspace, Gemini and its developer platform. Users who dictate long text, transcribe recorded audio or work with specialized terminology could benefit if the model performs as described. Who's exposed: Standalone transcription and dictation tools may face more pressure where Google bundles comparable functionality into Chrome, Workspace and Android workflows. It is too early to assess the competitive impact from the provided material alone.