Python API Reference¶

Full reference for the ModelDock Python SDK.

Import¶

import modeldock as md

Core Functions¶

load(name, backend=None, auto_install=None)¶

Auto-install if missing, then return a ready-to-use client.

Returns a runtime-native client (e.g., ollama.Client).

install(name, backend=None)¶

Explicit download of a model.

Returns ModelRef.

install_category(category, backend=None)¶

Bulk install models by category.

Returns list[ModelRef].

update(name, backend=None, confirm=False)¶

Pull a newer tag. Destructive — removes then re-downloads. Requires confirm=True.

Returns ModelRef.

remove(name, backend=None)¶

Uninstall a model.

verify(name, backend=None)¶

Integrity check. Returns bool.

Discovery Functions¶

list()¶

Browse the full catalog. Returns list.

search(query)¶

Search by name, capability, or category. Returns list.

installed()¶

Models present locally. Returns list[ModelRef].

info(name)¶

Sizes, capabilities, variants. Returns model info object.

categories()¶

Available categories. Returns list[Category].

recommend(task)¶

Guided pick for a task. Returns list.

Execution¶

run(name, prompt=None, backend=None, **opts)¶

Run a completion or interactive REPL.

Cache¶

md.cache.status()    # list cache entries
md.cache.path()      # real cache directory
md.cache.clean()     # safe: removes only corrupt/partial entries
md.cache.clean(force=True)  # wipes everything

Configuration¶

md.configure(
    backend="ollama",
    auto_install=True,
    log_level="INFO",
    progress_style="rich",
    cache_dir="/custom/path",
    ollama_host="http://localhost:11434",
)

Explicit Manager¶

mgr = md.Manager(
    backend="ollama",
    auto_install=True,
    ollama_host="http://localhost:11434",
)

Exported Names¶

__all__ = [
    "load", "list", "search", "installed", "info", "categories",
    "recommend", "install", "install_category", "update", "remove",
    "verify", "run", "cache", "configure", "Manager", "ModelManager",
    "ModelRef", "RuntimeBackend", "Category",
]