Flux 1.1 Pro vs Ultra: When You Don't Need to Pay 1.5x

Two tiers of the same model, one priced at exactly 1.5 times the other, and the higher tier sitting right next to the lower one in every model picker. The default instinct is to take the expensive one. For most marketing image work that instinct buys something you will not use, and costs something you might.

Key Takeaways

• Pro: $0.032 per image. Ultra: $0.048 per image. Ultra is exactly 1.5x.
• Image-to-image costs the same as text-to-image on both tiers — no editing premium anywhere in this family.
• Ultra adds two ultra-wide aspect ratios: 21:9 and 9:21, for eleven in total against Pro's nine.
• Ultra drops webp. Pro outputs webp, jpg or png; Ultra outputs jpg or png only.
• Ultra is therefore not a superset of Pro. It trades one capability for another.

The price gap, exactly

Both tiers currently run at 20% off, and both charge the same rate whether the request is text-to-image or image-to-image.

The flat image-to-image rate is worth noting on its own. Several image families charge double for editing, so a Flux pipeline that leans on i2i refinement keeps a predictable per-image cost in a way that some cheaper-looking alternatives do not.

At a thousand images, the tiers separate by $16 — $32 against $48. That is not a large number in isolation. It becomes one at catalogue scale, and it is money spent on a capability that most briefs never call for.

What Ultra actually adds

The concrete, measurable addition is aspect ratio range. Ultra accepts 21:9 and 9:21; Pro does not.

Those are cinema-wide and its vertical mirror, and there are real jobs that need them:

• Full-bleed website hero bands where a 16:9 crop leaves letterboxing.
• Ultra-wide display advertising and billboard-style placements.
• Vertical companion crops for the same campaign, generated natively rather than cropped down.

If a brief specifies one of those, Ultra is the tier that can produce it without an awkward crop, and the 1.5x is simply the price of the format.

What Ultra takes away

Here is the part that rarely appears in a tier comparison: Ultra cannot output webp.

Pro returns webp, jpg or png. Ultra returns jpg or png. For anything delivered to a browser that is a real regression. webp typically lands well under an equivalent-quality jpg, and for a page carrying a dozen generated images that difference shows up in load time and in Core Web Vitals.

The practical consequence is a conversion step. A team that standardised on webp delivery and then upgrades to Ultra does not just pay 1.5x per image — it also adds a transcode stage at the end of a pipeline that previously had none, with the storage and cache-invalidation bookkeeping that comes with it.

A rule of thumb

• The brief specifies 21:9 or 9:21 → Ultra. Nothing else in the family produces it.
• Delivery format is webp → Pro, unless the ratio requirement overrides it.
• High volume, standard ratios → Pro. The premium buys a format range the job does not use.

Neither tier is a mistake. The mistake is assuming the higher number is strictly better, when the spec sheet says it is a trade.

Pro's everyday marketing workflow and Ultra's product photography work cover what each tier looks like in practice, and the full Flux family comparison puts both against the rest of the lineup.

Both endpoints are reachable through Siray.ai with one API key, so testing the actual trade takes a model-string change rather than a second integration.

Create your free Siray account and run the same brief through Pro and Ultra before standardising on either.