ModelRush: One API for text, image, video, and voice models

We built ModelRush for teams that want to evaluate and ship multimodal models without rewriting their integration for every provider. The public catalog shows callable models, supported API types, regional availability, and pricing from the same versioned registry used by execution and billing. You can start with an existing OpenAI SDK, change the base URL to https://api.modelrush.ai/v1, and select a live ModelRush model ID. Compatibility is limited to documented surfaces; not every model supports every operation or region. The docs cover request timeouts, bounded retries, asynchronous media jobs, uploads, and signed webhooks. We would especially value feedback on model discovery, regional selection, and the clarity of the request-tracing workflow. Current availability and pricing remain visible on the model and pricing pages rather than being frozen into this launch copy.