Most people use Antbase to route text. But the same OpenAI-compatible endpoint does more than chat — it also generates images and videos. There's no separate image API to learn, no second SDK to install, and no new base URL. You change the model string, and that's it.
Create where you already work
Because Antbase speaks the OpenAI API, image and video generation work anywhere that does. That means your own application, a quick script, an OpenAI-compatible client like Cursor, or the Antbase playground — all without leaving the stack you've already wired up. No tab switching, no copying prompts into a separate tool. You send a chat request; you get an image or a video back.
Two model strings: ant:image-gen and ant:video-gen
Antbase exposes generation as two virtual models. Call ant:image-gen for images and ant:video-gen for video, exactly like any other model. ANT picks the best underlying model for the request — or you can pin a specific one when you want a particular look.
from openai import OpenAI
client = OpenAI(base_url="https://antbase.ai/v1", api_key="ant-...")
# Images — ANT routes to a current image model behind one handle.
resp = client.chat.completions.create(
model="ant:image-gen",
messages=[{"role": "user", "content": "a paper-craft ant carrying a glowing orb, studio light"}],
)
# Video — same call, different handle.
video = client.chat.completions.create(
model="ant:video-gen",
messages=[{"role": "user", "content": "slow dolly over a neon city at night, cinematic"}],
)Let your assistant do the conversational part — drafting and refining the prompt — and let Antbase do the generation. Because it's the same endpoint as your text calls, the same API key, billing, and routing apply. Nothing new to provision.
See exactly how these route, and which models sit behind each handle, in the virtual-models docs:
Or just open them in the playground and generate something:
Your assistant handles the conversation. Antbase handles the generation — through the one API you're already calling.
