Status
System status
Live availability for every model on the API and in the Studio. Each model is health-checked continuously and requests are routed to healthy capacity automatically.
No models are currently available
We are bringing capacity online. Check back shortly.
Health checks starting
Page refreshed
Video models
- Seedance 2.0ByteDance · Text to Video, Image to Video, Reference to VideoUnavailableSuccess (24h)—p50 time—Last checked—
- Seedance 2.5ByteDance · Text to Video, Image to Video, Reference to VideoUnavailableSuccess (24h)—p50 time—Last checked—
- Seedance 2.0 FastByteDance · Text to Video, Image to Video, Reference to VideoUnavailableSuccess (24h)—p50 time—Last checked—
- Seedance 2.0 MiniByteDance · Text to Video, Image to Video, Reference to VideoUnavailableSuccess (24h)—p50 time—Last checked—
- Wan 3.0Alibaba · Text to Video, Image to Video, Reference to VideoUnavailableSuccess (24h)—p50 time—Last checked—
- Wan 3.0 PrimeAlibaba · Text to Video, Image to Video, Reference to VideoUnavailableSuccess (24h)—p50 time—Last checked—
- MiniMax H3MiniMax · Text to Video, Image to Video, Reference to VideoUnavailableSuccess (24h)—p50 time—Last checked—
- MiniMax H3 MaxMiniMax · Text to Video, Image to VideoUnavailableSuccess (24h)—p50 time—Last checked—
Image models
- Seedream 5.0 ProByteDance · Text to Image, Image EditUnavailableSuccess (24h)—p50 time—Last checked—
- Seedream 5.0 LiteByteDance · Text to Image, Image EditUnavailableSuccess (24h)—p50 time—Last checked—
- Seedream 5.0 FlashByteDance · Text to Image, Image EditUnavailableSuccess (24h)—p50 time—Last checked—
What the states mean
- Operational. Accepting and completing generations normally.
- Degraded performance. Working, but slower or with a higher failure rate than usual.
- Outage. Most generations are failing. Failed jobs are refunded automatically.
- Checking. Not enough recent traffic to judge yet — checks are in progress.
- Unavailable. Temporarily not accepting new generations.
How we measure
Every model is checked continuously against live capacity. Success (24h) is the share of generations that completed successfully in the last 24 hours; p50 time is the median time from submission to a finished output. When one capacity pool degrades, requests are routed to a healthy one automatically.
Building against the API? See error handling and the generation lifecycle. To report an incident, email support@cheapergen.com.