> For the complete documentation index, see [llms.txt](https://docs.telys.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.telys.ai/resources/faq.md).

# FAQ

Short, direct answers to the questions people ask most about Telys — privacy, offline use, GPUs and model downloads, bring-your-own embeddings, Windows, and pricing.

Quick answers to common questions. Each links to the page with the full story.

## Is my data private?

Yes. Telys is on-device: collections live in a local directory you choose, and your documents, vectors, and metadata do not leave the machine as a side effect of using Telys. `telys serve` makes **no outbound network connections**, and over TCP it refuses anonymous clients (an access token is required and compared in constant time). The only traffic Telys initiates is control-plane sign-in, license/key exchange, and the signed runtime download — never your memory contents. See [What leaves the machine](https://docs.telys.ai/security/what-leaves-the-machine.md) and [Data residency & privacy](https://docs.telys.ai/security/data-residency-privacy.md).

## Does it work offline?

Yes, after you sign in once. `telys login` does a device-code sign-in and installs a verified runtime and license; from then on, collection operations run locally with no network, and the license is verified **offline**. You need connectivity again only to sign in on a new device, renew a license, or download a runtime update. See [On-device & offline-first](https://docs.telys.ai/understand-telys/on-device-offline.md) and [Sign in & activate a license](https://docs.telys.ai/start-here/sign-in.md).

## Do I need a GPU or a model download?

No — not with the lexical embedders. `AlgentaBigramEmbedder` and `AlgentaMultigramEmbedder` hash character n-grams into a normalized vector deterministically, using the native kernel. No model weights to download and no accelerator required.

```python
from telys import Telys
from telys.embedding import AlgentaMultigramEmbedder

engine = Telys("./memory")
emb = AlgentaMultigramEmbedder()   # no GPU, no model download
col = engine.create_collection(
    "docs", dim=emb.profile.dimension, partition_by="scope",
    embedder=emb,
)
```

You need a GPU or downloaded weights only if you *choose* a dense model through bring-your-own embeddings. See [Lexical vs dense embeddings](https://docs.telys.ai/understand-telys/lexical-vs-dense.md) and [Choosing an embedder](https://docs.telys.ai/embeddings/choosing-an-embedder.md).

## Can I use my own embeddings?

Yes. `CallableEmbedder` wraps any function `list[str] -> [N, D] float32` behind an `EmbeddingProfile`, so you can plug in OpenAI, BGE / sentence-transformers, or any other model — dense or lexical.

```python
from telys.embedding import CallableEmbedder, EmbeddingProfile

profile = EmbeddingProfile(provider="openai", model_id="text-embedding-3-small",
                           model_version="1", dimension=1536)
embedder = CallableEmbedder(my_embed_fn, profile)   # my_embed_fn: list[str] -> [N, 1536]
```

The collection binds to that space via `profile.space_id()`, so reopening it re-attaches the right embedder. See [CallableEmbedder (bring your own)](https://docs.telys.ai/embeddings/callable-byo.md) and the [OpenAI embeddings recipe](https://docs.telys.ai/embeddings/recipe-openai.md).

## Does Telys run on Windows?

Through WSL2. The runtime installer supports `macos-arm64`, `linux-x86_64`, and `linux-arm64` (experimental); on Windows, install and run Telys inside a WSL2 Linux environment. Intel macOS is not yet supported. See [Supported platforms](https://docs.telys.ai/operations/supported-platforms.md).

## Which Python versions are supported?

`requires-python >= 3.10`. The package is pure-Python (`py3-none-any`) with classifiers through 3.13; because the runtime loads via `ctypes`, it also runs on 3.14. See [Python version support](https://docs.telys.ai/operations/python-version-support.md).

## Is it free? What does it cost?

There is a free tier. `telys_developer` is free for one device — sign in with `telys login`. Paid tiers are `telys_pro` (commercial), `telys_enterprise` (contact sales), and `telys_team` (self-host / serve). See [Plans & licensing](https://docs.telys.ai/licensing/tiers-and-pricing.md); current prices are on [telys.ai/pricing](https://telys.ai/pricing).

## How do I know my install is trustworthy?

Verify it. The runtime is signed and checked offline before it is trusted, and you can re-run the check any time:

```bash
telys runtime verify
```

See [Verify your install](https://docs.telys.ai/start-here/verify-install.md) and [Verify the runtime signature](https://docs.telys.ai/security/runtime-signature-verification.md).

## Next steps

* [Support & contact](https://docs.telys.ai/resources/support.md) — where to ask a question that isn't answered here.
* [What is Telys](https://docs.telys.ai/start-here/what-is-telys.md) — the full picture in one page.
* [Troubleshooting](https://docs.telys.ai/troubleshooting/overview.md) — fixes for common errors.
