Fish Speech targets Linux and an NVIDIA GPU, but Fish Audio S2 Pro runs on Apple Silicon through Murmur's MLX build, and Murmur also runs it on Intel Macs. See the specs, license terms, and steps.
Short answer: Fish Speech is Fish Audio's open text to speech project, and the model you can realistically run on a Mac is Fish Audio S2 Pro, which Murmur ships through an MLX bridge on Apple Silicon and also runs on Intel Macs. The official repository and its installation docs target Linux or WSL on an NVIDIA GPU or CPU, so on a Mac the practical route is a local app that has already converted the weights, not the repo's own commands. This article covers what S2-Pro is, the exact requirements Fish Audio publishes, how the app path works, how the two routes differ, and what the license actually permits.
Every model detail below comes from pages I opened on 2026-10-19: the Fish Speech GitHub README, the Fish Audio S2-Pro Hugging Face card, the installation and requirements page, and the two license texts. Murmur details come from the app's own model catalog. If a number is not on one of those pages, it is not here.
What Fish Speech is today
Fish Audio now ships two weights from the same repository: Drama 3 and S2-Pro. The README lists Drama 3 as a 35B-A3B mixture of experts (3B active per token) and calls it the new-generation flagship for multi-character scenes and long content. S2-Pro is the previous-generation flagship with faster inference. The README's model table puts S2-Pro at 4B parameters, and the Hugging Face card describes a Dual-Autoregressive design with a 4B slow autoregressive backbone plus a 400M fast autoregressive module, roughly 4.5B in total. Some MLX conversion cards round that up to 5B, which is the figure Murmur's catalog uses. Both models share the same command line and API server; the type is read from the checkpoint's config.json.
S2-Pro is the piece most people mean by "Fish Speech" in a local workflow. Per the Hugging Face card it was trained on over 10 million hours of audio across 80+ languages, and it accepts free-form inline [tag] instructions such as [whisper] or [laughing] rather than a fixed tag list, with more than 15,000 unique tags supported. It also does native multi-speaker generation by separating characters with speaker IDs, and, per the README, it reproduces a timbre from 10 to 30 seconds of reference audio. On a single NVIDIA H200 the card reports a real-time factor of 0.195, about 100 ms to first audio, and 3,000+ acoustic tokens per second.
One detail matters before you download anything: the code and the S2-Pro weights do not carry the same license. The repository notice states that the codebase and its associated weights are released under the Fish Audio Community License, while "the S2-Pro weights remain under the Fish Audio Research License." Those two documents grant very different commercial rights, and we will quote them directly later.
Can you run Fish Speech on a Mac?
From the repository, no, not in any documented way. The installation page has a requirements table: Drama 3 needs at least 48 GB of GPU memory (FP8) and runs on "NVIDIA GPU only: no CPU, Apple MPS or AMD ROCm, and no --compile." S2-Pro needs at least 24 GB and runs on an "NVIDIA GPU or CPU," under a system line that reads "Linux, WSL." A separate warning adds that the compile option "is not supported on Windows and macOS." There is no macOS or Apple MPS row anywhere in the requirements, so the repo path on a Mac is CPU-only at best and outside the documented target. The realistic Mac route is an MLX conversion, which is exactly what Murmur ships.
That leads to a simple decision test. Use the table below to pick a lane before you spend an evening on setup.
| Situation | Run Fish Speech from the repo | Use Fish Audio S2 Pro in Murmur |
|---|---|---|
| Your Mac is the only machine you have | Docs list Linux or WSL; no macOS or Apple MPS path for S2-Pro or Drama 3 | Runs on Apple Silicon and Intel Macs with macOS 15 or later, with MLX weights on Apple Silicon |
| You have a Linux box with an NVIDIA GPU | Documented target; S2-Pro asks for at least 24 GB of GPU memory, Drama 3 at least 48 GB (FP8) | Not required, though the same model can run locally on the Mac |
| You want to avoid Python and PyTorch setup | Install goes through conda or uv, Python 3.12, and a PyTorch index | No Python setup; install the model inside the app |
| You need the newest Fish Audio flagship | Drama 3 is the current flagship, but it is NVIDIA GPU only and absent from the Mac path | Murmur ships Fish Audio S2 Pro, not Drama 3 |
| You need the original Research License weights exactly as published | You download fishaudio/s2-pro yourself and manage the checkpoint | You run an MLX conversion of S2 Pro, which carries the same Fish Audio Research License and needs a separate written license from Fish Audio for commercial use |
How Fish Audio S2 Pro runs inside Murmur
Murmur's catalog lists Fish Audio S2 Pro as a "5B dual-AR model with voice cloning and expression tags," running through an MLX bridge rather than the repo's Python stack. It gives two download sizes: about 11 GB in bf16, and about 6.7 GB in an 8-bit variant that Murmur prefers on lower-memory Macs. Murmur's non-Kokoro models ask for at least 16 GB of unified memory, and the app keeps Kokoro available as the lightweight fallback. If you are on a 16 GB machine, the 16 GB Mac guide covers how the app chooses between variants.
The app exposes a specific set of languages for this model: English (US and UK), Japanese, Chinese, Korean, Spanish, Portuguese, Arabic, Russian, French, and German. The Fish Audio model card lists 80+ languages for the model itself, with Japanese, English, and Chinese as Tier 1 and Korean, Spanish, Portuguese, Arabic, Russian, French, and German as Tier 2. Treat the model card as the ceiling and the app's selector as what you can pick today.
The working steps are short:
- Install Fish Audio S2 Pro from the model list. Murmur downloads the 8-bit variant on a lower-memory Mac and the bf16 variant on a larger one.
- Add a reference voice. Fish cloning needs at least 3 seconds of audio, 10 to 30 seconds works best, and the app caps a sample at 210 seconds. The 8-bit MLX variant also requires the exact spoken transcript, so record with the words written down or transcribe the take first.
- Only clone a voice you own or have written permission to use. The Community License states a consent rule of its own, and it is the right habit under any license.
- Type the script and add expression tags such as [whisper] or [laughing] where the delivery should change. S2-Pro accepts free-form tags, and Murmur exposes the Fish controls plus speed control.
- Generate and export. Murmur writes standard audio files, so the result drops straight into an editor; the voice cloning guide walks through preparing a clean sample.
Run it from the repo or use a local app
| Task | Fish Speech from the repo | Fish Audio S2 Pro in Murmur |
|---|---|---|
| Operating system | Linux or WSL per the install docs; macOS is not listed | macOS 15 or later on Apple Silicon or Intel |
| Hardware for S2-Pro | NVIDIA GPU or CPU in the main path (the docs also list other accelerators); at least 24 GB of GPU memory recommended | Apple Silicon; 16 GB of unified memory minimum, with an 8-bit variant preferred on lower-memory Macs |
| Setup | Python 3.12, conda or uv, then hf download fishaudio/s2-pro | Install the model inside the app |
| Voice cloning input | 10 to 30 seconds of reference audio plus its transcript | A saved sample from 3 seconds (10 to 30 seconds best), plus the transcript for the 8-bit variant |
| Expression control | Free-form [tag] instructions, 15,000+ tags | Fish expression tags and speed control |
| Multi-speaker | Native speaker IDs inside one text | Check the app for project-level multi-speaker workflows |
| Streaming | Streaming server; the README publishes production numbers on H200 GPUs | Batch generation and export, not a live stream |
| Licensing to review | Fish Audio Community License for code, Research License for S2-Pro weights | Fish Audio's Research License travels with the S2-Pro weights, including the MLX conversions, and commercial use needs a separate written license from Fish Audio |
| Cost | Free to download; your own hardware and time | $49 one-time, plus a 7-day refund; Fish Audio S2 Pro and cloning need Pro or the 3-day trial, and the free plan is Kokoro only |
The honest read: pick the repo when you have an NVIDIA Linux box, need the original checkpoints, or want to fine-tune and serve through SGLang. Pick the app when your only hardware is a Mac and you want the model running today. The voice studio versus DIY comparison covers the broader trade-off between maintaining a Python stack and using a packaged app.
The license decides whether you can ship the audio
Two licenses sit on this project, and they are not interchangeable. The Fish Speech source code and the Drama 3 weights fall under the Fish Audio Community License, version 1.0, last updated October 13, 2026. Section III(a) grants a free commercial license "only while Your Annual Revenue is less than USD 500,000." Once your annual revenue reaches or passes that threshold, you must notify Fish Audio in writing within 30 days, the community license continues for a 90-day transition, and after that any commercial use needs "a separate written license agreement from Fish Audio." Section III(c) is stricter: offering a Hosted Service, meaning third parties generate speech from inputs they supply, always requires a separate license regardless of your revenue.
The S2-Pro weights are different. The Hugging Face card marks them under the Fish Audio Research License, last updated March 7, 2026, and the card states plainly: "Research and non-commercial use is permitted free of charge. Commercial use requires a separate license from Fish Audio." The license text is blunter still: "Any use of the Fish Audio Materials or Derivative Works for a Commercial Purpose requires a separate written license agreement from Fish Audio. No commercial rights are granted under this Agreement." So the $500,000 revenue allowance in the Community License applies to the code and the Drama 3 weights, not to the S2-Pro weights you run through an MLX build.
The two documents share some terms. You own the outputs you generate "to the extent permitted by applicable law," and if you redistribute the materials or ship a product that uses them, you must show the "Built with Fish Audio" attribution. The Community License forbids generating speech that imitates the voice of an identifiable natural person without that person's prior consent. The Research License instead requires you to follow Fish Audio's Acceptable Use Policy, which I could not retrieve, so confirm the consent terms there before you rely on either. If you need commercial rights, the license names the address, business@fish.audio.
Practically, a Murmur purchase licenses the app; it does not relicense Fish Audio's weights. If you plan to sell the audio you generate, read the license that ships with the exact weights you use, and contact Fish Audio if you need a commercial agreement. The commercial license comparison and the open model license checklist are useful for auditing a project.
What can go wrong
- The repo has no documented macOS or Apple MPS path. The requirements table lists Linux and WSL, and the docs state the compile option is not supported on Windows or macOS.
- Drama 3 is NVIDIA only: the docs exclude CPU, Apple MPS, and AMD ROCm, and it needs at least 48 GB of GPU memory in FP8.
- Disk and memory planning: S2-Pro is about 11 GB in bf16 or about 6.7 GB in 8-bit inside Murmur, and the repo pulls the main weights plus a codec checkpoint. Make room before you start.
- Python and dependency friction: the repo installs Python 3.12 through conda or uv, and the docs explicitly note pyaudio install errors that may need conda to fix.
- The 8-bit MLX variant needs the exact reference transcript, so a clean recording with known words saves a retry.
- License mismatch: the code and the S2-Pro weights are governed by different licenses, so a permission that applies to one does not automatically apply to the other.
- The docs say Drama 3 switches to float16 on older GPUs without bfloat16 and has not been tested on real hardware there, and for S2-Pro you may need to pass the --half option yourself.
Frequently asked questions
Can I run Fish Speech on a Mac?
Not through the official repository in any documented way. Its installation requirements list Linux and WSL, with an NVIDIA GPU or CPU, and there is no macOS or Apple MPS entry. Fish Audio S2 Pro does run on Apple Silicon through Murmur, which converts the model to MLX, and Murmur also runs it on Intel Macs. Drama 3 is NVIDIA only and is not part of Murmur's catalog.
Is Fish Speech free, and can I use it commercially?
The code and Drama 3 weights are under the Fish Audio Community License, which grants free commercial use only while your annual revenue is below USD 500,000 and requires a separate license above that, with hosted services always needing one. The S2-Pro weights are under the Fish Audio Research License, which permits research and non-commercial use free of charge and grants no commercial rights. Check the license for the exact weights you use.
How much reference audio does S2-Pro need for voice cloning?
The Fish Speech README says 10 to 30 seconds of reference audio is enough to reproduce a timbre. In Murmur, a Fish voice sample can start at 3 seconds, 10 to 30 seconds is recommended, and a sample is capped at 210 seconds. The 8-bit variant also needs the exact spoken transcript.
Which languages does Murmur's Fish Audio S2 Pro support?
The app's catalog lists English (US and UK), Japanese, Chinese, Korean, Spanish, Portuguese, Arabic, Russian, French, and German. The Fish Audio model card lists 80+ languages for the model itself, so treat the card as the ceiling and the app's selector as what you can pick right now.
Does Murmur offer a free trial?
Murmur has a free plan with a Kokoro voice and scripts up to 1,500 characters, and a one-click 3-day Pro trial that the app starts itself, one per Mac. Fish Audio S2 Pro and voice cloning are Pro features, so you need the trial or a license to try them. The website edition is $49 one-time, and purchases have a 7-day refund policy.
Do I need an NVIDIA GPU to use S2-Pro?
Not in Murmur, which runs Fish Audio S2 Pro through an MLX bridge on Apple Silicon and also runs it on Intel Macs. Running the model from Fish Audio's own repository does need specific hardware: the docs list an NVIDIA GPU or CPU for S2-Pro, and an NVIDIA GPU only for Drama 3.
Sources
- Fish Audio: Fish Speech GitHub repository README (models, architecture, license notice)
- Fish Audio: S2-Pro model card on Hugging Face (parameters, tags, languages, license)
- Fish Audio: Installation and requirements (GPU memory, CPU/MPS support, weights download)
- Fish Audio: Inference documentation (command line, API server)
- Fish Audio Community License, version 1.0 (code and Drama 3 weights)
- Fish Audio Research License (S2-Pro weights)
Try Murmur - $49 one-time. No subscriptions, no cloud, no per-character fees.
Originally published at murmurtts.com
