Skip to content

Models

Jump to bottom Edit New page
Ari Mayer edited this page Mar 23, 2026 · 7 revisions

Models

Browse 4,000+ AI models across 85+ providers. The Models tab provides a three-column layout with providers, model list, and detail panel. The CLI offers interactive pickers and JSON output for scripting.

Models tab

TUI

Layout

The Models tab uses a three-column percentage layout:

Column Width Content
Left 20% Provider list with model counts
Center 45% Model list with capability indicators
Right 35% Detail panel for the selected model

Use Tab or arrow keys to move focus between columns.

Capability indicators (RTFO)

Each model in the list shows four capability indicators:

Letter Meaning Active color
R Reasoning Cyan
T Tool use Yellow
F File/attachment support Magenta
O / C Open weights / Closed weights Green / Red

Inactive capabilities show as · in dark gray.

Provider categories

Providers are grouped into five categories:

Category Description Examples
Origin Model creators OpenAI, Anthropic, Google
Cloud Cloud platform providers AWS Bedrock, Azure, GCP Vertex
Inference Inference-focused providers Together AI, Groq, Cerebras
Gateway API gateway/routing services OpenRouter, Helicone
Dev Tool Developer tools and IDEs Cursor, Vercel

Filters and sort

Key Action
s Cycle sort (name, date, cost, context)
S Toggle sort direction (ascending/descending)
1 Toggle reasoning filter
2 Toggle tools filter
3 Toggle open weights filter
4 Toggle free models filter
5 Cycle provider category filter (All, Origin, Cloud, Inference, Gateway, Tool)
6 Toggle category grouping in the provider panel

Active filters are shown in the panel title. Sort direction is indicated by ↓ (descending) or ↑ (ascending) followed by the sort field name.

Detail panel

The detail panel shows full information for the selected model:

  • Header -- model name and provider
  • Capabilities -- reasoning, tool use, attachments, open weights
  • Modalities -- input and output modality support (text, image, audio, video, pdf)
  • Pricing -- input, output, cache read, and cache write costs per million tokens. Free models show "Free" in green
  • Limits -- context window size and maximum output tokens
  • Metadata -- release date, last updated, knowledge cutoff, provider documentation URL

Missing values display as an em-dash.

Copy and open

Key Action
c Copy model-id to clipboard
C Copy provider/model-id (full reference)
o Open provider docs in browser
D Copy provider docs URL
A Copy provider API URL

Search

Press / to search across model names, IDs, and providers. The search query appears in the panel title as [/query].

CLI

CLI models list

Interactive model picker

models list
models list anthropic    # Prefilter by provider

Opens an inline terminal picker with a side-by-side layout (model table + detail preview). Inside the picker:

  • / starts a live text filter over model ID, name, and provider
  • s cycles sort modes
  • S reverses the current sort
  • c copies the selected model ID to clipboard
  • Enter prints the selected model's full details

Show model details

models show claude-opus-4-5-20251101
models show gpt-4o

Prints a formatted detail view with ID, provider, family, limits, pricing, capabilities, and metadata.

If the query matches multiple providers or variants in an interactive terminal, the CLI reopens the picker with the matching candidates so you can choose the exact model.

Search models

models search "gpt-4"
models search "claude opus"

Uses the same matcher and interactive picker flow as models list.

List providers

models providers
models providers --json

JSON output

All model commands support --json for scripting:

models list --json
models show claude-opus-4-5 --json
models search "llama" --json
models providers --json

Repository · Issues · Releases · brew install models · MIT License

models wiki

Features

Reference

Development

Clone this wiki locally