Realtime generation status

Generation states

After submitting a prompt, Lovision runs the generation task and shows its status near the prompt input or on the generated layer.

State	What it means
Generating	The AI model is actively producing the result. Progress may be shown as a percentage or a spinner.
Completed	The result is ready. The generated output is placed in your project automatically.
Failed	The task did not complete. An error message usually explains why.
Canceled	You or the system stopped the task before it finished.

Why tasks fail

Common reasons for a failed generation:

Invalid prompt — the prompt contained content that the model declined to process. Revise the prompt and try again.
Reference too large — an attached file or reference image exceeded the model's input size limit. Resize or compress the reference.
Timeout — the model took too long to respond. This can happen with large outputs or complex requests. Adjust the prompt or model and submit a new generation.
Insufficient credits — the generation consumed the last of your available credits mid-task. Add credits and resubmit.
Model unavailable — the selected model is temporarily offline. Switch to an alternative model in the model selector.

Handling a failed task

Read the visible error message on the generation status.
Adjust the prompt, reference files, or model.
Submit a new generation from the prompt input.

Canceling a running task

Click the × or Cancel button next to the in-progress task to stop it. Canceled tasks do not consume credits (or refund them, depending on the model — see Credits and concurrency limits).

Tips

If a task consistently fails with the same prompt and model, try a different model to rule out a model-specific issue.
Long video or multi-frame generation tasks take more time than image tasks — the Generating state for video may last several minutes.
Keep the browser tab active during generation; some operations rely on an active connection to receive the result.