FLUX.2 pro vs Nano Banana Pro API: request specs side by side
Request parameters that differ between black-forest-labs/flux.2-pro and google/nano-banana-pro on Sume: resolution tiers, aspect ratios, references and price.

On Sume, black-forest-labs/flux.2-pro lists no resolution tier, while google/nano-banana-pro takes a resolution of 512, 1K, 2K, 4K and native aspect ratios only. Both take input_references and n. This is a spec table, not a ranking.
Sume's side is read from the Image API docs, Image 1.0 and catalog code; each vendor's own limits are on its page, read 2026-09-29.
How do the two requests differ?
The main split is the resolution tier. Nano Banana Pro lists tiers; the FLUX.2 rows list none. Custom pixel sizes are a separate matter: POST /v1/images does not serve explicit pixels in v1, and Image 1.0's docs say image_size custom pixels go to models such as Flux, while on Nano Banana a size maps to the native aspect_ratio.
| Item | `black-forest-labs/flux.2-pro` | `google/nano-banana-pro` |
|---|---|---|
aspect_ratio | 1:1, 16:9, 9:16, 4:3, 3:4, 4:5, 5:4, 3:2, 2:3, 21:9, 9:21, 1:2, 2:1 | auto, 21:9, 16:9, 3:2, 4:3, 5:4, 1:1, 4:5, 3:4, 2:3, 9:16 |
resolution tiers | None listed | 512, 1K, 2K, 4K |
Custom pixels via Image 1.0 image_size | Accepted | Maps to a native ratio |
input_references | Up to 10 | Up to 10 |
| One image | $0.04 | $0.19 |
What do the vendors themselves limit?
Black Forest Labs' editing page says FLUX.2 takes up to 8 reference images through its API (10 in its playground) and outputs up to 4MP. Google's docs list up to 5 character images and up to 6 object images on Nano Banana Pro (Gemini 3 Pro Image). These are the vendors' own APIs; on Sume the ceiling is the catalog's input_references descriptor, 10 for both.
Which parameters break a request?
A parameter the chosen model does not list returns 400 unsupported_parameter. Sending resolution: "4K" to a model with no tier list is the usual case, so read the descriptors first.
curl "https://api.sume.com/v1/images/models/black-forest-labs/flux.2-pro/endpoints" \
-H "Authorization: Bearer $SUME_API_KEY"How do I run the same prompt on both?
Send two requests that differ in model and drop fields one model does not list. Each returns Sume-hosted URLs under data[].url.
for M in black-forest-labs/flux.2-pro google/nano-banana-pro; do
curl -s -X POST https://api.sume.com/v1/images \
-H "Authorization: Bearer $SUME_API_KEY" \
-H "Content-Type: application/json" \
-d "{\"model\": \"$M\", \"prompt\": \"A red kettle on a wooden table\", \"aspect_ratio\": \"4:5\"}"
doneSources
Related posts
More in Models
- gemini-omni-flash-preview shuts down September 30: what to change
Google's deprecations list gemini-omni-flash-preview with a 2026-09-30 shutdown and gemini-omni-1.1-flash as its replacement. What changes, and the Sume id.
- Gemini Omni Flash vs Veo 3.1: which does Google recommend when
Google's video page names Gemini Omni Flash the default, Veo 3.1 for scene extension, last-frame control or legacy pipelines. Facts side by side, plus Sume ids.
- Does Sume Auto use GPT Image 2.5? What sume/auto picks for images
Sume's docs say Auto image routing continues to use Flare, GPT Image 2.5. Sume never says which family ran on a call, so pin openai/gpt-image-2.5 if it matters.
- GPT Image 2.5 Flare vs Sunburst: what each model page says
OpenAI calls Flare its model for everyday image generation and Sunburst its most capable for generation and editing. What each page lists, and both ids on Sume.
Written by Sume