Models¶

List available Codex models and their capabilities.

Listing Models¶

models = client.models.list()

for m in models:
    print(f"{m.slug:30} context={m.context_window}")

Model Fields¶

Field	Type	Description
`slug`	`str`	Model identifier (e.g. `"gpt-5.1-codex-mini"`)
`display_name`	`str`	Human-readable name
`context_window`	`int \\| None`	Maximum context length in tokens
`reasoning_levels`	`list[str]`	Supported reasoning effort levels
`input_modalities`	`list[str]`	Supported input types (e.g. `["text", "image"]`)
`supports_parallel_tool_calls`	`bool`	Whether parallel tool calls are supported
`priority`	`int`	Display priority

Caching¶

Model results are cached for 5 minutes to avoid redundant API calls. Force a refresh:

# Uses cache if available
models = client.models.list()

# Bypasses cache
models = client.models.list(force_refresh=True)

Finding a Specific Model¶

models = client.models.list()

# Find by slug
mini = next((m for m in models if m.slug == "gpt-5.1-codex-mini"), None)
if mini:
    print(f"Context window: {mini.context_window}")
    print(f"Reasoning levels: {mini.reasoning_levels}")