Models
Live video and image models, inputs, availability, and Token pricing
Call GET /v1/models before sending production traffic. It is the source of
truth for model IDs, inputs, sizes, durations, prices, and availability.
Video models
| Model ID | Input | Duration | Price |
|---|---|---|---|
seedance-2.0 | Text, image | 4–30s | 210 / 5s |
seedance-2.0-fast | Text, image | 4–30s | 169 / 5s |
seedance-2.0-mini | Text, image | 4–30s | 105 / 5s |
wan-2.7-text-to-video | Text | 2–15s | 100 / second |
wan-2.7-image-to-video | Image | 2–15s | 100 / second |
wan-3.0-video | Text, image | 2–30s | 100 / second |
Image models
qwen-image-3.0-pro is available for text-to-image and reference-image
generation. It costs 42 Tokens for 1K output, 79 Tokens for 2K output, plus 4
Tokens per reference image.
The upstream image catalog is also protocol-integrated: gpt-image-2,
gemini-3.1-flash-image-preview, gemini-3-pro-image-preview,
doubao-seedream-5.0-lite, doubao-seedream-4.5,
doubao-seedream-5.0-pro, wan2.7-image, wan2.7-image-pro,
z-image/turbo, z-image/turbo-lora, z-image-turbo/image-to-image,
z-image-turbo/image-to-image-lora, youchuan/text-to-image-v7, and
youchuan/text-to-image-v81.
Models whose catalog entry has availability: "pricing_required" are not yet
billable and reject generation requests with HTTP 503. This prevents an
unverified upstream price from causing an incorrect debit.
curl https://api.mistrounter.ai/v1/models \
-H "Authorization: Bearer $MISTROUNTER_API_KEY"See Generate an image, Wan video, and Seedance for request examples.