Releasesby GlyphDrama 3
Drama 3 Text-to-speech model you direct in plain language. Describe tone, pacing, and character without a fixed set of audio tags. Use model ID drama-3 in the API. Drama 3 guide | Text to Speech API
Also shipped Oct 14, 2026Fish Audio in October 2026 →
Workspace-level package repositories for Databricks Apps are generally available
Workspace admins can now configure default Python and npm package repositories for Databricks Apps. Apps inherit the workspace configuration while still allowing app-level overrides. See Configure default package repositories .
The Apache Hive catalog endpoint in the Lakehouse runtime catalog is now generally available (GA)
The Apache Hive catalog endpoint in the Lakehouse runtime catalog is now generally available (GA). This integration lets you use the metastore as a serverless, central metadata layer, enabling seamless data sharing and unified querying across Apache Spark, Apache Hive, and BigQuery.
Metadata migration from Dataproc Metastore to Lakehouse is now generally available (GA)
Metadata migration from Dataproc Metastore to Lakehouse is now generally available (GA). You can perform a one-time backfill migration or dry run to migrate metadata from a Dataproc Metastore service to an Apache Iceberg REST catalog endpoint or a Hive catalog endpoint using the Google Cloud console, the gcloud CLI, or the REST API.
Conversational analytics now lets you create a conversation with all of the data sources in multiple...
Conversational analytics now lets you create a conversation with all of the data sources in multiple projects, or create a conversation with a dataset . These features are generally available (GA).
You can publish a BigQuery Conversational Analytics agent in Gemini Enterprise
You can publish a BigQuery Conversational Analytics agent in Gemini Enterprise . This feature is now generally available (GA).
The AI.AGG function now supports the output_schema argument, which lets you specify the schema of the...
The AI.AGG function now supports the output_schema argument, which lets you specify the schema of the function output as a comma-separated list of fields. This feature is generally available (GA).
More from Fish AudioEvery Fish Audio release →
Fish Audio S1 retirement
Fish Audio S1 retirement S1 is deprecated and will be retired on December 31, 2026. After that date, requests that specify s1 are served by s2.1-pro and billed as s2.1-pro , at the same price. To migrate, switch the model ID to s2.1-pro and convert (parenthesis) emotion tags to [bracket] tags. Migrate from S1 | Model deprecations
TTS language field
TTS language field Text-to-speech requests accept an optional language field, a BCP-47 tag of the text such as en , es-MX , or zh-TW . It selects the rules used to spell out numbers, dates, units, and symbols before synthesis, and the language used for word timestamps. It does not change the voice, accent, or audio quality.
More normalization languages
More normalization languages Text normalization now has rules for Turkish ( tr ) and Tagalog ( tl , fil ), and Portuguese follows Brazilian conventions, so pt-BR is supported too. Without language , Turkish and Tagalog text is detected, and Chinese written mostly in Traditional characters uses the Traditional Chinese rules. Supported normalization languages
Transcribe 1 Pro
Transcribe 1 Pro Speech-to-text model for short clips and long recordings, including multi-speaker conversations. Transcripts can include inline speaker markers, speaker turns, and emotion and vocal-event cues. Use model ID transcribe-1-pro on POST /v1/asr Speech to Text | Models overview
Fish Audio S2.1 Pro
Fish Audio S2.1 Pro Production text-to-speech model that improves on S2-Pro in quality, latency, and throughput. Supports 83 languages on one model, with the same [bracket] natural-language control and multi-speaker dialogue as S2. Use model ID s2.1-pro for production workloads that need TTFA and DPA guarantees.
Fish Audio S2
Fish Audio S2 Next-generation text-to-speech model with inline emotion cues, multi-speaker dialogue support, and 80+ languages. S2 introduces [bracket] syntax for natural language control over emotion and paralinguistic cues (e.g., [whisper] , [laugh] , [emphasis] ).
Fish Audio S1
Fish Audio S1 Historic rebrand from Fish Speech to Fish Audio. #1 ranking on TTS-Arena2 with industry-leading performance. S1 (4B params): 0.008 WER, 0.004 CER - Available on Fish Audio Playground S1-mini (0.5B params): 0.011 WER, 0.005 CER - Open source on Hugging Face 64+ emotional expressions with RLHF integration and multilingual support for English...