Models in the app
Download and install models
From Explore to Stored: formats, fit guidance, the download queue, installation status and recovery.
Noema separates transferring files from checking and preparing them. A model or voice is ready only after its required installation steps finish.
On this page
What can be downloaded
Noema discovers and downloads compatible GGUF, MLX, ExecuTorch, and Core ML model packages. Apple Foundation Models are supplied by the operating system and appear when supported; they are not normal model downloads.
Read the model page before downloading
- Compatibility and estimated RAM fit for the current device.
- Quantization, file size, architecture, and context guidance.
- Vision, tool-calling, reasoning, or audio capability badges.
- Required projector, tokenizer, or other companion dependencies.
- Publisher, source repository, license, and provenance notes.
From Explore to Stored
- Open Explore → Models and search or browse the catalog.
- Choose a model, then select a compatible file or quantization.
- Review fit guidance and dependencies before tapping Download.
- Monitor progress. Interrupted supported downloads can resume rather than restart from zero.
- Open Stored when the model is ready, adjust its settings, and load it.
Queue and model-loading status
Use Pause All and Resume All to manage the download queue. Noema checks available storage before large downloads. Model loading shows named preparation steps, can be cancelled, and offers Retry with an explanation when loading fails.
Follow the current device memory recommendation. Estimates account for memory-mapped models on iOS; a model’s file size and its configured context still affect whether it fits.
Understand installation progress
| Status | Meaning |
|---|---|
| Downloading | Required files are being transferred. |
| Paused or waiting for connectivity | The transfer is waiting for an action or connection. |
| Verifying | Noema is checking downloaded files. |
| Preparing | Files are present, but additional preparation is running. |
| Ready | The installation is available to use. |
| Failed or Incomplete Download | The installation needs attention. |
Finishing a transfer does not always mean preparation has finished, especially for neural voice models.
Recover a model download
- Open the download list.
- Find the interrupted download.
- Choose Resume for a paused transfer or Retry for a failed attempt, when available.
- Wait for verification and installation to finish.
- Open the model in Stored and load it.
Supported downloads can reuse saved transfer progress. Recovery depends on the source and the state of the downloaded files.
Use Pause when you intend to continue later. Stop Download removes partially downloaded data after confirmation.
Recover a neural voice installation
Open Settings → Voice Model and check the installation status.
- Choose Retry for an incomplete installation.
- Choose Repair when it is offered.
- Wait until the model reports Ready.
If the downloaded files are intact, Noema can retry preparation in place. If files are missing or invalid, recovery may require downloading replacements.
Background behavior
On Mac, downloads continue while Noema is running and resume after the next launch if the app was quit. On iPhone and iPad, background transfers depend on system availability; force-quitting stops automatic continuation.
Troubleshooting downloads
- Free enough storage for the final file and temporary transfer data.
- Keep the source license and gated-repository requirements in mind.
- If a vision model loads without image support, check for its required projector.
- If a file is marked incompatible, choose a supported architecture or runtime instead of forcing it.
- Off-grid Mode blocks new model and dependency downloads.