TokenstromAll about AI Toolbox

Phonon: Gradium's tiny text-to-speech models now run locally in Ollama

Gradium has released its Phonon text-to-speech family in the Ollama library. At 31 to 40 million parameters, the models are very small and turn text straight into spoken audio. They only run on Apple Silicon, though.

Model releases Try it yourself Oct 22, 2026 3 min read Source: Ollama-Bibliothek
Illustrative image, AI-generated

What it's about

Gradium has made Phonon, its family of text-to-speech models (speech synthesis: text is turned into spoken language), available in the Ollama library. According to the description, Phonon produces audio as it generates. That makes it useful for reading an assistant's replies aloud, narrating text passages or giving local apps a voice. Ollama is a program that lets you run AI models on your own computer. At the time of writing, the model page showed 54 downloads.

Variants and languages

By default, Ollama pulls "Phonon 31M English". The number stands for about 31 million parameters, the values the model learned during training. That's very few for a speech model. This model ships with a single voice called "default".

If you need another language, pick a 40M model. The source lists checkpoints (saved training states of the model) for English, French, German, Spanish and Portuguese, plus a multilingual checkpoint. The excerpt doesn't give the exact tags for each variant; you'll find them on the model page.

Licenses

Licenses differ by variant:

Variant License Use
31M English Apache 2.0 commercial use allowed
40M (EN, FR, DE, ES, PT, multilingual) CC BY-NC 4.0 non-commercial only

If you want to use the German model, you can only do so for non-commercial purposes.

What you need to try it

  • Hardware: According to the source, the Ollama packages only run on Apple Silicon (Macs with an M chip). Windows, Linux and Nvidia GPUs aren't mentioned, so they aren't supported according to this source.
  • Software: Ollama 0.40.9 or newer.
  • Resources: With 31 to 40 million parameters, memory requirements are low. The source gives no exact figures for RAM or disk space.
  • Command: ollama run phonon "Good morning. Here is a short update before your first meeting."

Practical notes

Phonon outputs 24 kHz mono audio. Your input text should contain only spoken content, because formatting characters such as Markdown headings can be read aloud, according to the source. For saving audio and using the speech API, Ollama points to its "Speech guide".

Takeaway

Phonon shows that speech output is possible locally with very small models, without a cloud service. The restriction to Apple Silicon and the non-commercial license for all multilingual variants limit the audience. The source says nothing about actual voice quality, latency or comparisons with other models; you can only judge that by trying it yourself. The default, with just one voice, is likely more of a starting point for many applications.

Sources

#text-to-speech#ollama#gradium#apple-silicon#open-weights
Read the original source

Try it yourself: Phonon 31M English

Explained step by step, no prior knowledge needed – on your own computer or free in your browser.

What can I do with it?

Phonon reads written text aloud with a synthetic voice. You can use it to voice a local assistant's replies, have texts read to you, or give your own apps speech output, all without the cloud.

Who is it worth it for?

Interesting for Mac users on Apple Silicon who want to test local speech output or build it into their own apps. According to the source, it isn't suitable for Windows or Linux users right now.

Size0.031 B parameters
Downloadunder 0.1 GB (about 1 min at 100 Mbit/s)
LicenseApache 2.0 (31M English); CC BY-NC 4.0 (40M-Modelle)
TypeSpeech output (text to speech)
Ollamaphonon

What your computer needs

Graphics card (NVIDIA) at least 1 GB VRAMalmost any graphics card with 4 GB or more (e.g. GTX 1650, RTX 3050)
Mac (Apple Silicon M1–M5) from 8 GB unified memoryThe GPU and CPU share the memory – the model can use about 70% of it
Storage approx. 2 GB freeModel download plus programs
Google Colab runs for freeThe free Tesla T4 GPU with 15 GB is enough

Values estimated from the model size – the Model Card may have more precise figures.

Which variant is right for me?

Many models come in several sizes or "quantizations." These are compressed versions: smaller and faster, but slightly less accurate. Rule of thumb: pick the largest variant whose memory requirement fits in your VRAM (video memory). The scripts below choose this for you automatically.

VariantDownloadNeeds approx.Compatible graphics cards
4-bit (Q4) Default under 0.1 GB1 GBalmost any graphics card with 4 GB or more (e.g. GTX 1650, RTX 3050)
8-bit (Q8) under 0.1 GB1 GBalmost any graphics card with 4 GB or more (e.g. GTX 1650, RTX 3050)
16-bit (Original) 0.1 GB1.5 GBalmost any graphics card with 4 GB or more (e.g. GTX 1650, RTX 3050)

The default 31M English model (Apache 2.0) is the simplest choice for English, including commercial use. For German or multilingual use, pick a 40M model, which may only be used non-commercially.

  • Runs only on Apple Silicon (Mac with an M chip), according to the source.
  • Requires Ollama 0.40.9 or newer.
  • At 31–40M parameters the model is very small, so memory needs should be low.
  • Output: 24 kHz mono audio.
How do I find out what my computer has?
  • Windows: Press Ctrl + Shift + Esc → Task Manager → Performance tab → click GPU on the left. At the bottom you'll see Dedicated GPU memory (that's your VRAM). You'll find system memory under Memory.
  • Mac: Apple menu at the top left → About This Mac. You'll see your chip (e.g. M2) and memory (e.g. 16 GB). On Apple chips, graphics and processor share this memory.
  • Linux: Type nvidia-smi in the terminal (shows your graphics card and its memory) and free -h for system memory.

Guide

Google Colab (in your browser, no installation)

Google Colab is a free service from Google. It gives you a computer with a graphics card in your browser for a few hours. All you need is a Google account. For this model: runs for free – The free Tesla T4 GPU with 15 GB is enough.

  1. Download the notebook. Click the button and save the file.

    Download colab.ipynb

  2. Open Colab. Go to colab.research.google.com and sign in with your Google account.
  3. Upload the notebook. In the window that appears, click Upload on the left (or choose File → Upload notebook at the top) and select the file colab.ipynb.
  4. Turn on the GPU. In the top menu, go to Runtime → Change runtime type → under Hardware accelerator, select T4 GPU → Save.
  5. Start everything. Runtime → Run all (keyboard shortcut Ctrl + F9). If you see the warning “This notebook was not authored by Google,” click Run anyway.
  6. Wait. The cells run one after another. A green checkmark appears to the left of each cell when it's done. Installation and download take about 4 minutes here.
  7. Try it out. The result appears below the last cell. To test something else, change the text in the cell and run it again with Shift + Enter. The last cell starts a proper chat with an input field.
  8. Clean up. When you're done: Runtime → Disconnect and delete runtime. This saves your free quota.

Good to know: The free GPU is limited (usually a few hours per day). After about 90 minutes of inactivity, Colab disconnects, and you'll have to start everything over.

Windows 10 / 11

The easiest way is Ollama, a free program that launches AI models with a single command.

  1. Update your graphics driver (NVIDIA cards only): Install the NVIDIA app or the latest driver. For AMD cards, use the latest Adrenalin driver.
  2. Install Ollama. Download the file OllamaSetup.exe from ollama.com/download/windows, double-click it, and choose Install. Afterward, a small llama icon appears in the taskbar at the bottom right.
  3. Open a terminal window. Press the Windows key, type PowerShell and press Enter. A blue or black window opens.
  4. Start the model. Copy this command ("Copy" button), paste it into the window with a right-click, and press Enter: 
    ollama run phonon
    The first time, Ollama downloads the model (under 0.1 GB (about 1 min at 100 Mbit/s)). After that, it's instant.
  5. Start typing. Once >>> Send a message appears, type your question and press Enter. Type /bye to end the chat.

Not enough memory? Pick a smaller variant from the table above, e.g. ollama run phonon.

Prefer automatic? A script that picks the right variant for you
  1. Download both files into the same folder: try_windows.ps1 demo.py
  2. Open the folder in File Explorer, right-click an empty spot and choose Open in Terminal.
  3. Paste this command and press Enter:
    powershell -ExecutionPolicy Bypass -File try_windows.ps1

If needed, the script installs Ollama via winget, measures your VRAM (video memory) and downloads the matching variant.

macOS

Macs with an Apple chip (M1 to M5) are ideal. Older Intel Macs are very slow.

  1. Install Ollama. Download it from ollama.com/download/mac, open the file, and drag Ollama into your Applications folder. Launch Ollama once – a llama appears in the menu bar at the top.
  2. Open Terminal. Press Cmd + Space, type Terminal, and press Enter.
  3. Start the model. Paste the command (Cmd + V) and press Enter: 
    ollama run phonon
    The first time, the model is downloaded (under 0.1 GB (about 1 min at 100 Mbit/s)).
  4. Start typing. Type your question and press Enter. Quit with /bye.
Pick the right variant automatically

try_mac.sh

cd ~/Downloads
bash try_mac.sh

Linux

  1. Check your GPU. nvidia-smi should show your card. If it doesn't, install the driver on Ubuntu and restart: 
    sudo ubuntu-drivers install
  2. Install Ollama: 
    curl -fsSL https://ollama.com/install.sh | sh
  3. Start the model: 
    ollama run phonon
    Exit with /bye. Use ollama ps to check whether the model is running on the GPU.
  4. Or let it pick the right variant automatically:

    try_linux.sh

    bash try_linux.sh

Try it

Good morning. Here is a short update before your first meeting.
Once upon a time, there was a little dragon who was afraid of fire.
Your package has been delivered and is waiting at your front door.

If something goes wrong

Update Ollama if startup fails (0.40.9 at minimum).

Markdown characters such as headings may be read aloud, so enter plain spoken text only.

For German, choose a 40M model; the default model is English only.

The 40M models are for non-commercial use only (CC BY-NC 4.0).

Saving audio and the speech API are covered in Ollama's Speech guide.

“out of memory,” a crash, or your computer freezes

The model is too big for your VRAM (video memory). Pick a smaller variant from the table above and close other programs, especially games and browsers with lots of tabs.

It runs, but extremely slowly

Your CPU is probably doing the work instead of your graphics card. In Ollama, the PROCESSOR column of ollama ps shows how much runs on the GPU. If it says "100% CPU", update your graphics driver or pick a smaller variant that fits entirely into VRAM (video memory).

“ollama” or “python” is not found

Close the terminal window and open it again so it recognizes the newly installed program. For Python on Windows: reinstall it and check Add python.exe to PATH.

Windows: "Running scripts is disabled on this system"

Run the script with exactly the command above: powershell -ExecutionPolicy Bypass -File try_windows.ps1. This allows execution for this one time only.

Error message when loading the model (version, "trust_remote_code", unknown architecture)

Brand-new models often need the latest software version. With Ollama: update Ollama. With Python: pip install -U transformers in the activated environment. The Model Card usually lists the exact version.

Tokenstrom generates the scripts and notebook automatically. Please take a quick look before running them – they're plain, readable text. Mind the model's license: Apache 2.0 (31M English); CC BY-NC 4.0 (40M-Modelle).

Ask the AI assistant

Questions about this article, installation or background? Our assistant knows this page and runs on our own machine, not in someone else's cloud.

Connecting …

The Assistant is an AI (Gemma 4) and can make mistakes – please check important details against the source. Questions are used only to generate an answer and are deleted after one hour. Please don't enter any personal data.

Comments

What do you think? Already tried it? Write it here, no sign-up needed.

  1. Loading comments …

Comments are public. Links aren't allowed, and we delete abusive posts. You can delete your own comment at any time in this browser. Learn more in the Privacy Policy.

More articles

Model releases Mercury 3: Diffusion language model writes an Asteroids game in 5 seconds Model releases Mistral Large 4 "Le Chonk": a 1-trillion-parameter model with open weights Model releases Kakao open-sources two Kanana-2 embedding models, built for Korean search