Clearer model prices in the bot

We redesigned the model screen in the Telegram bot so you can compare available models and their prices before connecting a project, without scanning a long list of technical identifiers.

What changed

• Models are grouped by family, making it easier to find a model line.
• Input, output, cache read, and cache write prices are shown separately when that usage type is available.
• Each entry includes its endpoint, usage format, and catalog refresh time.

Improvements

• Long-table line wrapping and price labels were corrected.
• Models from different families no longer get mixed, and the last catalog refresh is visible.

Compatibility and next steps

No API key or client change is required. Before configuring a project, check that the model is available to your key in the bot or with GET /v1/models.

See the bot guide for the current way to open the model list.