FLUX.2 max grounding search: what it is, what Sume lists

FLUX.2 [max] is the only FLUX.2 tier that searches the web when prompted, so it can ground an image in real-time information such as sports scores or weather. Sume's image catalog lists black-forest-labs/flux.2-pro and black-forest-labs/flux.2-flex, not max, so there is no FLUX grounding search on Sume.

BFL facts are from its text-to-image page and overview; Sume facts from Image generation, read 2026-10-01.

Which FLUX.2 tiers have grounding search?

The BFL overview table answers it directly.

What does Sume list?

Two FLUX.2 ids: black-forest-labs/flux.2-pro and black-forest-labs/flux.2-flex. Per the vendor table neither does grounding search. The docs name GET /v1/images/models as the list of what is served, so read it before you plan around any model id.

Can I send a search instruction in the prompt?

A Sume image prompt is a plain text description of the desired image. Writing "look up today's score" into it does not give a model without search a way to look anything up. If your image depends on live facts, fetch them yourself and put them in the prompt text. For another model's grounding story, see Nano Banana grounding on Sume.

What should I do instead?

Resolve the facts in your own code, then call flux.2-pro or flux.2-flex with the finished text. That keeps the output reproducible from the prompt you logged.