Repository Public
1<br />23Gemini 3.5 Transcribe is a speech-to-text model based on Gemini's audio understanding capabilities. It provides low-latency, accurate transcription with utterance-based language detection, speaker diarization, word-level timestamps, Smart transcription, and custom vocabulary speech biasing.4[Try in Google AI Studio](https://aistudio.google.com/prompts/new_chat?model=gemini-3.5-transcribe)56## Documentation78Gemini 3.5 Transcribe converts speech to text from audio files. It automatically detects 85+ languages, handles multi-language code-switching, identifies individual speakers, outputs word-level timestamps, and adapts to domain-specific terms through custom vocabulary biasing.910Visit the [Audio transcription](https://ai.google.dev/gemini-api/docs/transcribe) guide for full coverage of non-streaming features. For real-time, low-latency streaming transcription over WebSockets, visit the [Live transcription](https://ai.google.dev/gemini-api/docs/live-api/live-transcribe) guide using `gemini-3.5-transcribe-live`. For detailed pricing, see the [Pricing page](https://ai.google.dev/gemini-api/docs/pricing#gemini-3.5-transcribe).1112## gemini-3.5-transcribe1314| Property | Description |15|---|---|16| Model code | `gemini-3.5-transcribe` |17| Supported data types | **Inputs** Audio (up to 1 hour) **Output** Text, Word annotations |18| Audio limits | **Unary audio limit** Up to 1 hour per request (up to 30 minutes when speaker diarization or word-level timestamps are enabled) |19| Capabilities | **[Audio transcription](https://ai.google.dev/gemini-api/docs/transcribe)** Supported **[Speaker diarization](https://ai.google.dev/gemini-api/docs/transcribe#speaker-diarization)** Supported **[Word-level timestamps](https://ai.google.dev/gemini-api/docs/transcribe#word-timestamps)** Supported **[Custom vocabulary](https://ai.google.dev/gemini-api/docs/transcribe#custom-vocabulary)** Supported **[Smart transcription](https://ai.google.dev/gemini-api/docs/transcribe#transcription-modes)** Supported **[Caching](https://ai.google.dev/gemini-api/docs/caching)** Not supported **[Code execution](https://ai.google.dev/gemini-api/docs/code-execution)** Not supported **[File search](https://ai.google.dev/gemini-api/docs/file-search)** Not supported **[Function calling](https://ai.google.dev/gemini-api/docs/function-calling)** Not supported **[Image generation](https://ai.google.dev/gemini-api/docs/image-generation)** Not supported **[Thinking](https://ai.google.dev/gemini-api/docs/thinking)** Not supported |20| Consumption options | **[Batch API](https://ai.google.dev/gemini-api/docs/batch-api)** Not supported **[Flex inference](https://ai.google.dev/gemini-api/docs/flex-inference)** Not supported **[Priority inference](https://ai.google.dev/gemini-api/docs/priority-inference)** Not supported |21| Versions | Read the [model version patterns](https://ai.google.dev/gemini-api/docs/models/gemini#model-versions) for more details. - `gemini-3.5-transcribe` (Unary) - `gemini-3.5-transcribe-live` (Live API) |22| Latest update | August 2026 |2324## Feature support and limitations2526Gemini 3.5 Transcribe supports the following features across its two endpoints:2728| Feature | Live Streaming (`gemini-3.5-transcribe-live`) | Audio File Processing (`gemini-3.5-transcribe`) | Notes / Limitations |29|---|---|---|---|30| **Language Auto-detection** | Supported (85+ languages) | Supported (85+ languages) | Includes mid-session code-mixing. |31| **Word-level Timestamps** | Not Supported | Supported | Degrades transcription accuracy. Incompatible with custom vocabulary. |32| **Custom Vocabulary Biasing** | Supported (up to 1,000 terms) | Supported (up to 1,000 terms) | Customers typically see best results with up to 100 terms. Incompatible with diarization and timestamps. |33| **Smart Dictation \& Formatting** | Supported | Supported | Includes filler word removal and intent-aware alphanumeric formatting. |34| **Speaker Diarization** | Not Supported | Supported (up to 8 speakers) | Attribution for 3+ speakers is experimental. Incompatible with custom vocabulary. |35| **Max Audio Duration** | 10 minutes per session. | Up to 1 hour | File processing is limited to 30 minutes when features like diarization or timestamps are enabled. |3637## Supported languages3839The following languages and BCP-47 language codes are supported for Gemini 3.5 Transcribe:4041| Language | BCP-47 Code | Language | BCP-47 Code |42|---|---|---|---|43| Afrikaans | `af-ZA` | Japanese | `ja-JP` |44| Amharic | `am-ET` | Javanese | `jv-ID` |45| Arabic (Egypt) | `ar-EG` | Kabuverdianu | `kea-CV` |46| Armenian | `hy-AM` | Kannada | `kn-IN` |47| Assamese | `as-IN` | Kazakh | `kk-KZ` |48| Azerbaijani | `az-AZ` | Korean | `ko-KR` |49| Belarusian | `be-BY` | Kyrgyz | `ky-KG` |50| Bengali (Bangladesh) | `bn-BD` | Latvian | `lv-LV` |51| Bengali (India) | `bn-IN` | Lingala | `ln-CD` |52| Bosnian | `bs-BA` | Lithuanian | `lt-LT` |53| Bulgarian | `bg-BG` | Macedonian | `mk-MK` |54| Bulgarian (Aromanian) | `rup-BG` | Malay | `ms-MY` |55| Burmese | `my-MM` | Malayalam | `ml-IN` |56| Cantonese (Traditional) | `yue-Hant-HK` | Maltese | `mt-MT` |57| Catalan | `ca-ES` | Mandarin Chinese (Simplified) | `cmn-Hans-CN` |58| Cebuano | `ceb` | Marathi | `mr-IN` |59| Central Khmer | `km-KH` | Mongolian | `mn-MN` |60| Croatian | `hr-HR` | Nepali | `ne-NP` |61| Czech | `cs-CZ` | Norwegian | `nb-NO` |62| Danish | `da-DK` | Oriya | `or-IN` |63| Dutch | `nl-NL` | Polish | `pl-PL` |64| English (Great Britain) | `en-GB` | Portuguese (Brazil) | `pt-BR` |65| English (India) | `en-IN` | Portuguese (Portugal) | `pt-PT` |66| English (United States) | `en-US` | Punjabi | `pa-IN` |67| Estonian | `et-EE` | Punjabi (Gurmukhi script) | `pa-Guru-IN` |68| Farsi | `fa-IR` | Romanian | `ro-RO` |69| Filipino | `fil-PH` | Russian | `ru-RU` |70| Finnish | `fi-FI` | Serbian | `sr-RS` |71| French | `fr-FR` | Sindhi (Arabic script) | `sd-Arab-IN` |72| Galician | `gl-ES` | Slovak | `sk-SK` |73| Georgian | `ka-GE` | Slovenian | `sl-SI` |74| German | `de-DE` | Spanish (Latin America) | `es-419` |75| Greek | `el-GR` | Spanish (United States) | `es-US` |76| Gujarati | `gu-IN` | Swahili (Kenya) | `sw-KE` |77| Hausa | `ha-NG` | Swedish | `sv-SE` |78| Hebrew | `he-IL` | Tajik | `tg-TJ` |79| Hindi | `hi-IN` | Telugu | `te-IN` |80| Hungarian | `hu-HU` | Thai | `th-TH` |81| Icelandic | `is-IS` | Turkish | `tr-TR` |82| Indian English | `en-IN` | Ukrainian | `uk-UA` |83| Indonesian | `id-ID` | Uzbek | `uz-UZ` |84| Italian | `it-IT` | Vietnamese | `vi-VN` |