AI Gateway terminates a streaming request after ~785–790 seconds with gateway_stream_timeout, although the calling Workflow step has Extended Max Duration enabled and is deployed with maxDuration: 1800.
The provider started streaming after 6 seconds and continued returning reasoning tokens, so this is not a time-to-first-token timeout.
Is there an undocumented end-to-end stream limit? Can it be increased, or is there an official workaround that keeps AI Gateway?
VERCEL_ENABLE_WORKFLOW_EXTENDED_MAX_DURATION=1 is set for this deployment.
Similar unresolved report: https://community.vercel.com/t/deepseek-v3-2-long-reasoning-timeout-on-vercel-ai-gateway/29956
