Will the default API Gateway timeout support a long LLM response?
Not necessarily. Google documents a 15-second default deadline for HTTP streaming such as SSE and chunked responses, with a configurable maximum of 3,600 seconds. The backend can impose a shorter timeout, so both layers must be configured and tested.
Answered in
Google Cloud API Gateway Can Stream AI Responses, but the Timeout Is Your ProblemGoogle API Gateway adds SSE, WebSocket and gRPC streaming in preview; the gateway’s deadline and backend timeout still determine whether a stream survives.
Read the full analysisOther questions this article answers
More developer tools questions
- What is the largest change in google-genai 3.0?
- Does the new API version default affect Vertex AI clients?
- Should I pin the SDK version?
- What is GitHub’s Copilot code review addressed rate?
- Does a high addressed rate mean Copilot reviews are accurate?
- Who can access the new repository-level report?
- What is GitHub’s new ACCESSIBILITY.md support?
- Does adding ACCESSIBILITY.md prove that a project conforms to WCAG?
Every answer on Crashtech is written by the editor of the article it comes from — never auto-summarised. Browse all answers or the Developer Tools beat.