Skip to main content

AI generation problems

What each generation error means, whether retrying will help, and how to get a better scope out of a weak one.

Who this is for: Anyone whose scope failed to generate or came back thin.

Generation errors

MessageCauseDoes retrying help?
AI is taking too long. Please try again.The request passed the 30-second limit for a single attempt.Yes, usually. A shorter brief also helps.
AI is busy right now. Please retry in a moment.The model provider was rate limited or overloaded. Byteflare already retried up to twice with increasing delays before showing you this.Yes, but leave a minute rather than hammering the button.
AI returned an unexpected response. Please try again.The response could not be parsed into a valid scope, even after automatic repair.Yes. If it recurs on the same brief, change the brief's wording — something in it is producing malformed output.
AI generation failed. Please try again.An unclassified failure.Retry once. If it persists, report it with the project and the time.

The scope generated, but it is thin

  1. Answer the clarifying questions properly

    They exist because the answers are the highest-value input. Skipping them produces generic output.

  2. Put the constraints in the brief

    Budget ceiling, immovable dates, what the client already has, and what is explicitly out of scope.

  3. Set the project type and complexity accurately

    These change the shape of the generated document, not just a label on it.

  4. Regenerate a single section rather than the whole scope

    Targeted regeneration keeps the parts you already accepted.

  5. Edit it

    The draft is a starting point. Your judgement about this client is not in the brief.

The pricing looks wrong

Check your hourly rate in Settings → Account. In a Team workspace an owner or admin may have enabled a locked organisation rate, which overrides personal rates for every scope generated in that workspace. See Admin and workspace management.