Model Management

This page covers the shared model store, canonical model IDs, manifests, and the model-management commands.

Public surface

• mere.run model list
• mere.run model capabilities
• mere.run model info
• mere.run model pull
• mere.run model remove
• mere.run model repair-manifests
• mere.run status
• mere.run setup

Default model store

By default:

~/Library/Application Support/MereRun/models

Override with:

export MERERUN_MODELS_DIR=/path/to/models

or:

swift run mere.run --models-root /path/to/models model list

Canonical model IDs

Examples:

• images: image-klein-nano, image-zimage-nano, image-klein-max, image-zimage-max
• text: text-chat-gemma4, text-chat-q35, text-chat-q35-nano, text-agent-deepseek-v4-flash, text-agent-qwen35-9b, text-code-qwen3, text-embed-qwen3-0.6b
• speech: speech-tts-qwen3-nano, speech-asr-parakeet
• vision: vision-ocr-lighton
• music: music-acestep
• video: video-ltx-av

The public runtime resolves these IDs directly, so docs and examples should use the canonical names shown by mere.run model list.

Runtime entrypoints

• Sources/MereRunCore/MereRunModelPaths.swift
• Sources/MereRunCore/MereRunModelManifest.swift
• Sources/MereRunCore/ModelResolver.swift

Command responsibilities

mere.run model list

Shows the canonical managed model table and installed status.

mere.run status

Combines the model inventory with a local API probe. It reports the active model-store path/source, installed managed models, whether the configured API server answers /health, and the model IDs returned by /v1/models.

mere.run model capabilities

Summarizes the current Mac, the managed models it can run, the preferred setup-agent tier, and cross-modality starter coverage. Pass --all to include models that are blocked by Apple Silicon or unified-memory requirements.

mere.run model info

Shows the resolved local install for one canonical model ID.

mere.run model pull

Downloads a managed model from its cataloged Hugging Face source. Pulls are checked against the managed capability catalog before download so low-memory machines do not fetch models they cannot run. Pass --allow-unsupported only when you intentionally accept that risk or are using external hardware.

mere.run setup

Guided onboarding for the shared model store and first local agent. The command offers a Pi-powered Mere agent, a BYOA prompt for Claude/Codex, or manual commands. The small local agent model is text-agent-qwen35-9b; hardware-tier setup can select Q35 nano, Qwen3-Coder Next, or DeepSeek V4 Flash. On 96 GB+ Apple Silicon Macs, text-agent-deepseek-v4-flash is the preferred managed setup-agent tier; Q35/Qwen agent models are alternatives, not upgrades.

mere.run model remove

Removes a managed install from the local store.

mere.run model repair-manifests

Repairs manifest metadata in the local store when that metadata is missing or stale.

Related docs

• Model Sources
• Configuration
• Testing Guide