ollama pull <model>agi models scanLocal model serversollama running http://localhost:11434 models: - <model>agi --provider ollama --model <model>
What local mode costs you.
- WhereThe CLI. The CLI is coming soon.
- AccountNot required.
- Cost$0.00. The model runs on hardware you already own.
Each surface reaches local models its own way.
agi models scan
The CLI probes Ollama and LM Studio on loopback, prints every installed model beside its base URL, and blocks any address that is not loopback before a request is built.
Managed Cloud only
The Electron Desktop contract does not accept local-inference commands or provider keys. Local mode belongs to the CLI.
Local mode on the phone
Local mode on the phone is built to keep the conversation, memory, and files on the device.
You point AGI at a server you already run.
Nothing is installed on your behalf. Start the runtime, pull or load a model the way you normally would, then let AGI discover what is there and name the provider explicitly. The CLI is coming soon.
ollama pull <model>Pull a model into a running Ollama server.agi models scanList local model servers found on loopback, and the models each one reports.agi --provider ollama --model <model>Start a session pinned to that runtime and model.agi --provider lmstudio --model <model>Start a session pinned to an LM Studio model.
The runtime checks the boundary before every request.
A session carries its own privacy authority, local, byok, or managed, written down beside its transcript, and the runtime re-reads it before each request it builds. When the selected model routes somewhere that authority does not cover, the send stops there and the error names the one command that would carry the work across properly.
That command forks a second session which starts with no history at all. It runs a secret scan over the messages you picked, shows you the exact payload with counts of what was included, excluded and truncated, and completes only when you send that reviewed draft back word for word. The local session it forked from keeps its own authority and its own transcript, unchanged.
The machine you own can run the model.
The CLI answers from Ollama and LM Studio on the machine you own. BYOK on the CLI is the reviewed way out on the days a local model is not enough; Desktop is managed-cloud only. The CLI is coming soon.