oterm/docs/installation.md
2026-10-03 15:41:10 +03:00

102 lines
2 KiB
Markdown

# Installation
!!! note
`oterm` works with local and hosted LLM providers. For local models, point it at [Ollama](https://github.com/ollama/ollama?tab=readme-ov-file#ollama), [vLLM](https://docs.vllm.ai/), [LM Studio](https://lmstudio.ai/), [llama.cpp](https://github.com/ggml-org/llama.cpp), or any OpenAI-compatible runner. For hosted providers (OpenAI, Anthropic, Groq, …), set the relevant API key. See [Providers and API keys](app_config.md#providers-and-api-keys) and the [`openaiCompatible`](app_config.md#openaicompatible-custom-openai-compatible-endpoints) config block.
=== "uvx"
```bash
uvx oterm
```
=== "brew (macOS)"
```bash
brew install oterm
```
=== "yay (Arch)"
Community package maintained by [Rafael Dominiquini](https://github.com/Dominiquini). May lag behind the latest release.
```bash
yay -S oterm
```
=== "nix-env (NixOS)"
Community package maintained by [Gaël James](https://github.com/gaelj). May lag behind the latest release.
```bash
nix-env -iA nixpkgs.oterm
```
=== "pip"
```bash
pip install oterm
```
=== "pkg (FreeBSD)"
Community package maintained by [Nicola Vitale](https://github.com/nivit). May lag behind the latest release.
```bash
pkg install misc/py-oterm
```
### Spoken responses
The [`speak`](tools/index.md#capabilities) capability needs the `speak` extra, which brings in `piper-tts` (GPL-3.0):
=== "uvx"
```bash
uvx "oterm[speak]"
```
=== "pip"
```bash
pip install "oterm[speak]"
```
## Updating oterm
Use the same package manager you installed with.
=== "uvx"
```bash
uvx oterm@latest
```
=== "brew (macOS)"
```bash
brew upgrade oterm
```
=== "yay (Arch)"
```bash
yay -Syu oterm
```
=== "nix-env (NixOS)"
```bash
nix-env -uA nixpkgs.oterm
```
=== "pip"
```bash
pip install --upgrade oterm
```
=== "pkg (FreeBSD)"
```bash
pkg upgrade misc/py-oterm
```