LiteLLM fixes ignored web-search cap that turned a limit of one into three searches and a 500

LiteLLM merged a pull request fixing a bug that ignored the 'max_agentic_loops' cap, after a live test showed a cap-one request making three Tavily searches and six OpenAI Responses calls before returning HTTP 500. With the fix, the same request made one search and two Responses calls, returned 200 with 'finish_reason: stop', and logged $0.0312 in key spend instead of $0.0755. LiteLLM is the AI gateway co-founded by Krrish Dholakia and Ishaan Jaffer.

LiteLLM fixes ignored web-search cap that turned a limit of one into three searches and a 500 In the PR's cap-one live test, LiteLLM went from three searches, six Responses calls and a 500 to one search, two calls and a stopped 200. By RuntimeWire Staff https://runtimewire.com/author/runtimewire-staff · Published An administrator of LiteLLM, the AI gateway co-founded by Krrish Dholakia and Ishaan Jaffer, could set 'max agentic loops: 1', ask for three searches, and watch the gateway do the expensive part anyway. In the live test reported in the merged pull request, the request made three Tavily searches and six OpenAI Responses calls before returning HTTP 500. After the fix, the same cap-one request made one search and two Responses calls, returned 200 with 'finish reason: stop', and logged $0.0312 in key spend instead of $0.0755.…