estudely-askai

Terminal CLI for querying Ollama models via the HTTP API.

Install

python -m venv .venv
source .venv/bin/activate
pip install -e .

Usage

askai --version
askai "what is the capital of france"
askai --models
askai --host http://localhost:11434 --model llama3.1 "hello"
askai --stream "stream me"

Options

--host      Ollama host (default: env OLLAMA_HOST or http://localhost:11434)
--model     Model name (default: env ASKAI_MODEL or llama3.1)
--timeout   Request timeout in seconds (default: 60)
--json      Print models as JSON (use with --models)
--stream    Stream response tokens
--models    List available models
--cloud     Use Ollama cloud host (requires OLLAMA_API_KEY)
--local     Force localhost host default

Cloud auth

Set OLLAMA_API_KEY to send Authorization: Bearer <key> on all requests.

Config file

Defaults can be set in ~/.config/estudely-askai/config.toml:

host = "http://localhost:11434"
model = "llama3.1"
timeout = 60