Docs · models
Models & seats
PriestAI checks model names in the order below. If a name is ambiguous, it lists the matches so you can choose. A seat is a role assigned to a model.
How model names are matched
Enter a model name or alias. PriestAI checks these options in order:
| Rung | Example | What is selected |
|---|---|---|
| Agent mode | openclaw, agent, priest, harness | PriestAI's tool loop on an available local model. These aliases select a local model. Cloud models selected by name can also use tools, subject to the same permission checks. |
| Installed local | llama-3.1-8b, or a unique prefix like llama | A model on this machine or one of your other machines, at the server where it measured best. An ambiguous prefix refuses and names every candidate. |
| Catalog id | anthropic/claude-sonnet-5 | A paid model over OpenRouter, checked against the live catalog first, so a typo cannot be billed. Case-insensitive: the id as the catalog spells it is the one that is billed. |
| Short name | claude, grok, astra | Matched against the catalog by id, suffix, and whole word — the word "astra" selects the one model whose name carries it, and a pricing variant of the same model counts as one model, not an ambiguity. A word that matches nothing gives an error naming what would work; your typo is never sent to a paid API. On the day a new model is listed, a launch that misses re-reads the live catalog once before refusing, to refresh a potentially stale cache. |
Where the model list comes from
The Models screen probes what actually answers on your machines: Ollama first (its default port, or the one its environment names), then any OpenAI-compatible runtime on the network — LM Studio, vLLM and the like. What answers is what is listed; a runtime that is down is shown as down, not skipped. /models in the console reads the same probe.
Putting a model on a machine
The panel's Discover tab and the console's /catalog read the same installable catalogue — the Ollama library and Hugging Face GGUF files, with real sizes. /pull llama3.1 downloads one onto this machine and shows the progress, or name a host to put it on any machine of yours that can download. Only Ollama can be told to fetch over HTTP; for other runtimes the console explains how models get there instead (LM Studio through its own Discover tab, for instance). /hosts lists every machine and whether it can download; /adopt adds a second computer running Ollama as a model server, checked at the moment you add it.
The default model
The model the panel starts with is set in a text file you can edit: upgrades\PRIEST-MODEL.txt. It is read on every turn — change it and the next message uses the new model, with no restart. The word auto is a rule, not a model name: it resolves to the best model that measured as capable across your machines right now.
Paid cloud models and the spending limit
A cloud model is used per call through OpenRouter with your key, and every call goes through the same spending check: the month's spending is compared, per step, against the limit you set in the Money panel, and each step's real cost — read from the response's own usage figures — is recorded per model. When the limit is reached, paid models stop, with the figures shown. Local inference has no provider fee and is not subject to the cloud spending limit. Cloud providers bill your account separately, and their billing may lag the figures shown in PriestAI.
The same tools for every model
Local and cloud models use the same tool handler and permission checks. Available tools depend on your configuration. Cloud requests are billed by the provider for each step.