Send a chat completion. Never name an LLM — omit model (the Intelliverse default routes at your plan tier) or pass intelliverse/air, intelliverse/pro, intelliverse/pro-max, intelliverse/fusion.

List the tier aliases your API key can call. These are the only model strings that exist.

LLM daily budget for the API key plus per-App-ID media credit wallets (image, video, voice, audio, audiobook, 3D).

Create an app (the entry point for knowledge bases, credit wallets, and per-app cost tracking). Returns the app id and a recommended tier based on the declared workload.

List apps in the workspace with their recommended tiers and routing profiles.

Mint a tier API key (air | pro | pro_max), optionally bound to an App ID and daily spend limit. The raw key is returned exactly once — store it.

List API keys (prefixes only) with tier, app binding, and spend limits.

Permanently delete an API key by id.

Add documents (raw text or URLs — URLs are scraped) to the app's pgvector knowledge base. Chunks are embedded and become searchable memory for that App ID.

Semantic search over the app's chunks. Returns matches with similarity scores.

RAG chat: retrieves relevant chunks for the App ID and answers with [doc:id] citations.

List documents ingested for an App ID.

Send transactional/newsletter email through the platform email engine (Pro plan+). Billed at the email.send10 SKU from the App ID's iv_credits wallet (hold → dispatch → settle); failed sends are never charged and partial batches settle on accepted recipients only.

Generate image | video | voiceover | audio (music/sfx) | gen3d assets. Credits are held from the App ID's wallet and settled on completion. Returns a job — poll with media_job_status.

Get status/output of a media generation job.