Who can use Google Pics? Plans, and an image API if you can't

Google Pics is rolling out to Google AI Pro and Ultra subscribers and to most Workspace business customers, according to Google's announcement. It lives at pics.new, works inside Docs and Slides straight away, and gets Drive integration in the weeks after the announcement. If your account is not in one of those groups yet, an image API is the other way to reach a Nano Banana-class model, and on Sume the entry requirement is an API key and a funded wallet instead of a Google plan.

Pics facts are from Google's post, read 2026-10-03. Sume facts are from the Image API docs.

Pics access, as announced

The page names three entry points: the standalone site, Docs and Slides, and (later) Drive. It names two eligible groups: AI Pro and Ultra subscribers, and most Workspace business customers. It does not give a per-country or per-edition list, so check your own admin console before assuming you have it.

What an API call needs

Sume's public Image API is POST /v1/images, and the catalog is GET /v1/images/models. The first request below lists every model with how many references and images per call it takes, so you can see what is on offer before you spend anything.

import os
import requests

key = os.environ["SUME_API_KEY"]
url = "https://api.sume.com/v1/images/models"
resp = requests.get(url, headers={"Authorization": f"Bearer {key}"}, timeout=30)
resp.raise_for_status()
for model in resp.json()["data"]:
    params = model["supported_parameters"]
    refs = params.get("input_references", {}).get("max", 0)
    n_max = params.get("n", {}).get("max", 1)
    print(model["id"], "refs:", refs, "n:", n_max)

Honest limits

An API is not Pics. You get generation and reference-guided edits, not a shared canvas or an object picker, and each call is billed per image; the per-model price is on the endpoints record, which the docs say already includes Sume's margin. Failed or cancelled generations are not billed.