@squidcloud/client
    Preparing search index...

    Variable BULK_INGESTION_SPEEDSConst

    BULK_INGESTION_SPEEDS: readonly ["economy", "auto", "fast"] = ...

    How quickly a bulk-ingestion job must finish, and so which provider tier Squid buys for it. The model, the prompts and the resulting chunks are identical at every speed — only the queue in front of them differs.

    • economy — provider batch, cheapest, tails up to 24 hours.
    • auto — Squid decides from the job's size. The default.
    • fast — runs inline, minutes rather than hours, at full list price.