Skip to content
← Models
Open weightsFish Audio/10 trustMetrics-checked
Drama 3 logo

Drama 3

Fish Audio's open 35B text-to-speech MoE for multi-character scenes — 16 speaker tags, voice cloning, sub-100ms streaming.

Visit official page →

About

Drama 3 is Fish Audio's text-to-speech model for multi-character scenes and long-form content: 35B total parameters with 3B active per token in a mixture-of-experts design. It supports up to 16 speaker tags per text, voice cloning, and natural-language direction of how a voice should sound, with streaming at under 100ms time-to-first-audio (provider-claimed). It launched with day-one SGLang-Omni support and an arXiv technical report. Released under the Fish Audio Community License, it is the biggest open TTS release of the week — #4 on Hugging Face trending with 6.16k downloads and 537 likes — and extends the open-weights wave from text and vision into controllable expressive voice.

Specs

Context window

Long-form TTS (document-scale input supported)

Pricing

Free weights; Fish Audio hosted playground has paid tiers

Free quota

Weights free to download; playground free tier available

API access

Self-host via SGLang-Omni; hosted playground at fish.audio

Weights

Open weights

Released

Not disclosed

Benchmarks

No public standardized TTS benchmark scores located; traction signal is #4 HF trending, 6.16k downloads, 537 likes within days of release.

Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.

Use cases

text-to-speechvoice-cloningaudiobooksopen-weights

Trust score — 5/10

/10 trustMetrics-checked

Scored 7/10 (METRICS tier, metrics-v1). Evidence: Fish Audio ships the widely used Fish Speech stack, the release includes a technical report and day-one inference-engine support, and #4 HF trending shows genuine demand. Docked for: (1) latency and quality claims are provider-reported with no standardized benchmark to check them against; (2) the Fish Audio Community License is a custom community license, not OSI-approved — commercial users must read it; (3) brand-new release, so maintenance history is unwritten. License: Fish Audio Community License (custom).

How the trust score is built

Full methodology →
Adoption · stars/downloads, log-scaled vs peers
8.0/10
Maintenance · recency, releases, issue load
7.0/10
Openness · license + docs
6.0/10
Manual review · sanity read + red-flag check
7.0/10
  • · Fish Audio ships widely used Fish Speech stack; arXiv technical report included
  • · Day-one SGLang-Omni support; #4 HF trending, 6.16k downloads, 537 likes
  • · manual review: no standardized TTS benchmark to verify quality/latency claims; custom license needs reading

Sources