models

NAME

models — list available LLM models and current selection

SYNOPSIS

models [options]

DESCRIPTION

models lists the LLM models available from your configured providers. Each model is displayed with pricing (input/output cost per million tokens), context window size, and capabilities. By default only the latest version of each model family is shown for the currently selected provider.

When online, benchmark data from Artificial Analysis is fetched and cached (24h TTL) to enrich output with intelligence scores and speed (tokens/sec). This can be skipped with --no-benchmarks.

The currently selected model is marked with ► in the output.

OPTIONS

• --all

  List models across all configured providers, not just the active one.

• --all-versions

  Show every model version. Without this flag, models are deduplicated by family (e.g. only the latest Claude Sonnet is shown).

• --provider <id>

  Restrict output to a specific provider (e.g. openai, anthropic).

• --json

  Emit structured JSON for programmatic consumption.

• --refresh

  Force re-fetch of benchmark data from Artificial Analysis, ignoring the cache.

• --no-benchmarks

  Skip benchmark enrichment entirely. Faster, works offline.

OUTPUT

Each row contains:

• Model ID — the canonical identifier (e.g. claude-sonnet-4-20250514)
• Pricing — input / output cost per million tokens
• Context window — displayed as K or M tokens
• IQ — intelligence index from Artificial Analysis (when available)
• Speed — median output tokens per second (when available)
• Reasoning — flag shown for models with extended thinking capability

Non-chat models (embeddings, TTS, image generation, etc.) are filtered out automatically.

PROVIDERS

Models come from providers — accounts you configure in SLICC (OpenAI, Anthropic, Google, xAI, Adobe, etc.). Each provider exposes its own model catalog. Use providers to manage them. If no accounts are configured, models exits with an error.

EXAMPLES

models                          # list models for current provider
models --all                    # list across every configured provider
models --provider anthropic     # only Anthropic models
models --all-versions           # include older dated versions
models --json                   # machine-readable output
models --no-benchmarks          # skip AA data, instant output

SEE ALSO

cost, providers