Which Nano Banana models can you call on Sume? Google lists five
Google's Gemini API lists five Nano Banana models; Sume's image catalog lists two, Nano Banana 2.1 and Pro. Nano Banana 2 ids still run, as 2.1.
Two of Google's five Nano Banana models are in Sume's image catalog: Nano Banana 2.1 and Nano Banana Pro. Nano Banana 2 is retired but still accepted, and it runs as 2.1. Nano Banana 2 Lite and the original Nano Banana are not listed on Sume's image catalog.
Google's lineup against Sume's
| Google model (API id) | On Sume? | Sume id |
|---|---|---|
| Nano Banana 2.1 (gemini-nano-banana-2.1) | Yes | google/nano-banana-2.1 |
| Nano Banana Pro (gemini-3-pro-image) | Yes | google/nano-banana-pro |
| Nano Banana 2 (gemini-3.1-flash-image) | Retired; runs as 2.1 | google/nano-banana-2 resolves to 2.1 |
| Nano Banana 2 Lite (gemini-3.1-flash-lite-image) | Not listed | None |
| Nano Banana (gemini-2.5-flash-image) | Not listed | None |
What Sume's two rows give you
| Field | Nano Banana 2.1 | Nano Banana Pro |
|---|---|---|
| Billed at 1K | $0.10 | $0.1875 |
| Billed at 4K | $0.20 | $0.375 |
| Resolution tiers in the row | 0.5K, 1K, 2K, 4K | 0.5K, 1K, 2K, 4K |
| Aspect ratios | 15, including 4:1, 1:4, 8:1, 1:8 | 11, up to 21:9 |
| Reference images | Yes | Yes |
Retired ids still work
If your code still sends google/nano-banana-2 or nano-banana-2, Sume runs it as Nano Banana 2.1 and stores the 2.1 id on the job. The catalog does not list the retired id. Update the string anyway, so your logs match the model that ran.
What this means for your pipeline
- You cannot select the cheapest Google tier (Lite) through Sume's Images API today. For the lowest cost, look at the 2.1 row at 0.5K ($0.075 billed) or other families such as Grok Imagine and Qwen Image at $0.025.
- Google states that all images generated by these models include a SynthID watermark (Google AI, read 2026-10-07). Keep that in mind when you publish.
- Sume model ids use
org/slug. Google's API ids are for calling Google directly and are not valid in a Sume request.
Picking between the two
Use 2.1 for volume: it is cheaper at every tier and has a 0.5K tier for drafts at $0.075. Use Pro when your prompts were tuned to it. Sume does not publish a quality ranking, so the decision is yours after a test on your own prompts.
Google documents more Nano Banana models than Sume lists. If you rely on a feature from Google's docs, check the endpoint's supported_parameters before you plan a workflow; the Sume descriptor is the rule that applies on Sume.
Check the live list
The catalog changes. Call GET /v1/images/models and filter by id instead of relying on a blog table. The Image API page lists the canonical ids and retired aliases.
Sources
Related posts
More in Models
- Native 4K or 2K image edits: pixel ceilings by model
How many pixels each model reaches: Flux 3 Image near 16.8 MP, GPT Image 2.5 capped at 8,294,400 on Sume, Ideogram 4.5 at 2K. What Sume serves.
- Gemini Omni 'no dialogue' prompt: steer the audio on Sume
Omni clips always carry sound. Google suggests 'no dialogue' and 'no extra sound effects' in the prompt. How to use that on Sume and what a test costs.
- Gemini Omni sound effects prompt: name each sound in 8 seconds
How to prompt footsteps, a door and breaking glass in a Gemini Omni clip: Google says describe audio explicitly. Sume request, timing and cost per take.
- Open-weights video news does not change your Sume bill: LTX, H3, Wan
LTX-2.5 and MiniMax H3 are reported open weights, but Sume still bills hosted per-second rows. Which ids to call and what each costs.
Written by Sume