The Open Weights
LatestModelsLeaderboardsCompanies
Subscribe
The Open Weights

The daily record of open-source AI. New model releases, leaderboards, and what's coming next — written for people who ship.

Refreshed every 12 hours

Discover

  • Latest releases
  • New today
  • Trending models

Browse

  • All models
  • Companies
  • Categories
  • Leaderboards

About

  • About
  • Editorial policy
  • Hardware estimates
  • RSS feed
  • llms.txt
  • Newsletter

© 2026 The Open Weights. An independent publication.

PrivacyTermsSMSAggregated by Claude · curated by humans.
LatestFishaudio3
FishaudioText → Speech

Fish Audio's Drama 3 brings MoE to open TTS

The open-source speech model pairs a 35B mixture-of-experts backbone with voice cloning and directed performance control.

Oct 1, 2026
NotableOther
Fish Audio Drama 3

Fish Audio has released Drama 3, an open-source text-to-speech model built on a mixture-of-experts architecture. The model carries 35 billion parameters in total but activates roughly 3 billion per inference pass, a design that aims to deliver the quality of a large model while keeping the compute cost of generation closer to that of a much smaller one.

Beyond raw synthesis, Drama 3 is pitched around two capabilities that matter for expressive speech: voice cloning and directed performance. The latter suggests finer control over how a line is delivered — tone, emotion, and pacing — rather than just converting text to flat narration. That makes it a candidate for use cases like audiobooks, dubbing, and interactive characters, where delivery is as important as intelligibility.

Why it matters

Mixture-of-experts has reshaped open large language models over the past two years, but it remains relatively uncommon in speech generation. Applying the approach to TTS is a notable bet that the same efficiency gains can carry over to audio.

  • Multilingual: the model lists Chinese, English, and Japanese among supported languages, with more indicated.
  • Efficient by design: only about 3B of 35B parameters are active per pass.
  • Open weights: the model is published on Hugging Face, though under a custom "other" license worth reviewing before commercial use.

As with any open release, the practical question is how Drama 3 holds up against both proprietary speech APIs and other open TTS efforts. For teams already building on open audio stacks, an efficient MoE model with explicit performance controls is a meaningful new option to evaluate.

Sources

  • fishaudio/drama-3

    Hugging Face

    Visit
OlderCactus Compute's Whistle brings speech-to-text to the edgeCactus Compute · Speech → Text · 2 months agoNewerCloudflare's Clef brings structured decisions to open modelsCloudflare · Vision-Language · last month

Get the model

Hugging Face

Specs

Parameters35B · MoE
Languageszh, en, ja +
Size39.6 GB
PrecisionFP8
ArchitectureFishOmniModel
LicenseOTHER
Downloads45.6K
Likes1.3K

Can you run it?

Needs a workstation GPU (48 GB) as published.

  • FP8 (as published)

    48 GB GPU (RTX 6000) or 2×24 GB · 64 GB Mac

    40.6 GB
Your machine
  • Mixture-of-experts: all 35.9B parameters must be resident even though only 3B are active per token — active count affects speed, not memory.

Based on Hugging Face weights metadata. Assumes an 8K context and ~1 GB runtime overhead; actual needs vary. How we estimate


Modalities

Text → Speech

The Weekly Weights

Every open release that mattered, one email a week.

0 comments

No comments yet. Be the first to weigh in.

More from Fishaudio

All Fishaudio releases →
Fish Audio S2-Pro
Fishaudio/Text → SpeechRuns on a laptop

Fish Audio's S2-Pro Brings Expressive TTS to Open Source

The new text-to-speech model can follow natural language instructions to control tone, clone voices from short clips, and speak multiple languages.

Mar 9, 2026

More in Text → Speech

All Text → Speech →
Haru-4B
Higurashi Ai/Any-to-AnyRuns on a laptop

Haru-4B Brings Full-Duplex Speech to Six Languages

A compact 4-billion-parameter model handles real-time, two-way voice conversation across six languages in a single system.

Nov 12, 2026
Zonos3-Mini
Zyphra/Text → SpeechRuns anywhere

Zyphra releases Zonos3-Mini multilingual TTS

The compact open text-to-speech model supports multiple languages and voice cloning from short audio samples.

Nov 9, 2026
OpenBMB/Text → Speech

OpenBMB's Phonon Brings High-Quality TTS On-Device

The new text-to-speech model promises benchmark-topping voice synthesis that runs locally rather than in the cloud.

Nov 5, 2026