Skip to main content

When using BYOK, let us determine WHEN we fallback on using the API key

Feature Request

Currently, BYOK functionality only lets us determine whether to prioritise using the API key first and t3 allowance afterwards, or our t3 usage allowance first then the API key afterwards. (And there’s the Custom setting that lets us select which models to use the API key for).

What if I want to prioritise using my base usage limit first, then fallback to the API key, then if that’s exhausted go into burst overage usage limit? Or some other order of priority, but the above was what I thought the Custom setting would let me do. Currently I’d need to manually switch over to prioritising API keys after I’ve run through my base usage limit.

Granted, I’d like this setup for image generation, which is kind of the only reason I plugged the Gemini API key into t3 chat, and API keys can’t be used for that currently.

Status: Closed1 comment

Log in to comment and vote

Comments1

  • Mark Florkowski

    Team

    What you’re describing (custom fallback chains like base → API key → burst) isn’t something we support today. The main reason is that it breaks the assumptions our pricing is built around—supporting that would shift usage in a way that increases our costs, and we’d need to raise prices to compensate. BYOK is really intended as an escape hatch for cases where your usage doesn’t fit within our plans, not as a primary way to optimize or route usage. For now, we keep it simple (API key first vs allowance first), but we may revisit this if there’s a way to support it without impacting pricing.