Command reference
This page lists the current developer-platform CLI surface from sumelabs/cli.
Unsupported older consumer-product commands are intentionally omitted. Do not
assume Image/Video/Music generators exist as CLI subcommands — they do not yet.
Discover exact schemas at runtime:
Auth and account
sume me and sume account get both read account context for the configured
key.
Read commands
sume models list remains as a deprecated alias for sume catalog list.
Jobs helpers
Assets
Asset registration and upload paths are write-gated. Prefer public HTTPS URLs in generation payloads when you do not need first-party asset ids. See Jobs and media.
Write and paid generation (Avatar)
Write commands require explicit confirmation. Paid Avatar generators are the only first-class generation submit path in the CLI today.
Use --confirm-paid when provider execution may spend credits. Use
--confirm-submit for non-paid writes such as job cancellation or asset
registration.
Avatar-video scripts must estimate to 4-60 seconds inclusive before submission.
Avatar Video quality defaults to plus when omitted.
Batch helpers also exist under sume avatars batch and
sume avatar-videos batch (plan / create / watch / result against local state
files). Inspect with sume avatars batch --help.
Not available as CLI generators
The CLI does not ship sume image, sume video, sume music, or similar
subcommands. Use the HTTP API instead:
| Family | Guide |
|---|---|
| Image 1.0 | Image |
| Video 1.0 | Video |
| Music 1.0 | Music |
After you submit those jobs via API, you can still recover them with
sume jobs status / sume jobs result / sume jobs download.
MCP (CLI vs hosted)
Local sume mcp reports coming_soon / launched: false in current releases.
sume mcp install is not available yet.
For Cursor, Claude, and other remote MCP clients, use the
hosted MCP endpoint at https://mcp.sume.com/mcp — that is a separate
surface from the local CLI command.
Machine-readable output
--agent redacts URL-like and sensitive account/workspace fields where the CLI
supports agent-safe output.
Skills
See Agent skills.