@mzwing/pi-model-info0.6.2
Completes third-party model metadata in Pi from the pi.dev and models.dev catalogs
Untested on omp
This is a pi package. No one has reported whether it works on omp yet. If you try it, add your result below.
Reports 0
No reports yet.
Readme
@mzwing/pi-model-info
A Pi extension that completes the metadata of third-party models — context window, max tokens, pricing, capabilities and thinking levels — from the pi.dev and models.dev catalogs, where Pi would otherwise use placeholders. It never creates providers, discovers models or writes to your files, and works alongside discovery extensions such as pi-openai-api-models-sync.
Install
pi install npm:@mzwing/pi-model-info
Usage
Nothing happens until a provider is opted in, in ~/.pi/agent/extensions/pi-model-info/config.json or .pi/extensions/pi-model-info/config.json (the project wins):
{
"providers": { "my-relay": {} }
}
A model is matched by its alias, then its exact id, then its id with at most one prefix and one suffix rule removed. An ambiguous match changes nothing, and fields you hand-wrote in models.json always win.
/model-info summarises what was completed, /model-info <provider>/<model> shows where each field came from, and /model-info refresh re-checks the catalogs.
Configuration
Per provider, under providers:
| Key | Default | Description |
|---|---|---|
catalogProvider |
— | Scope lookups to one catalog provider, e.g. openrouter |
costMultiplier |
1 |
Markup applied to catalog pricing |
costPolicy |
catalog |
zero forces free, keep leaves pricing alone |
contextWindowPolicy |
catalog |
min never raises Pi's limits, keep leaves them alone |
capabilityPolicy |
catalog |
widen only adds capabilities, keep leaves them alone |
useCatalogName |
false |
Rename models to their catalog names |
mapThinkingLevels |
false |
Derive thinking levels from models.dev |
allowDynamic |
false |
Silence the warning when a refreshing provider's list gets frozen |
models |
— | Per-model alias, override, skip, prefixes and suffixes |
Rules strip an affix before matching. -free and :free are built in and set the cost to zero:
{
"rules": [
{
"id": "mainfei",
"kind": "suffix",
"value": "-mainfei",
"override": { "cost": { "input": 0, "output": 0, "cacheRead": 0, "cacheWrite": 0 } }
}
]
}
See the example config and the JSON Schema for every option.