Why did my AI response stop early?

Sometimes an AI provider stops before finishing its answer. When it returns finish_reason: length, the response reached the provider's output limit.

This commonly happens when:

• Max Tokens is set too low in BoltAI.
• The provider ends generation early, which can sometimes happen while its servers are under heavy load.

Continue the response

BoltAI keeps the partial response and shows Continue generating at the bottom of the chat. Click it to send Continue in the same conversation, allowing the model to continue from where it stopped.

If it keeps happening

• Increase or reset Max Tokens in the chat, project, agent, or account settings where you customized it.
• Try again later or switch to another model or provider.
• Ask the model to produce the answer in shorter sections.