{"id":"ai-gateway","name":"AI Gateway (Azure APIM)","homepage":"https://aka.ms/ai-gateway/labs","repo_url":"https://github.com/Azure-Samples/AI-Gateway","category":"devops","subcategories":["api-gateway","azure","governance","load-balancing","observability"],"tags":["azure","api-management","apim","gateway","load-balancing","rate-limiting","semantic-caching","mcp","bicep","jupyter","finops"],"what_it_does":"A collection of hands-on labs demonstrating how to build enterprise AI gateways using Azure API Management, covering load balancing, rate limiting, semantic caching, MCP integration, and multi-model routing with 30+ Jupyter notebook labs.","use_cases":["Implementing centralized AI model governance with Azure API Management policies","Load balancing and rate limiting across multiple Azure OpenAI endpoints","Integrating MCP servers with enterprise security via OAuth and managed identities"],"not_for":["Non-Azure environments - deeply tied to Azure APIM and Azure OpenAI","Production-ready gateway deployment - these are educational labs, not a turnkey product","Simple local development setups without Azure subscriptions"],"best_when":"You are building enterprise AI infrastructure on Azure and need patterns for governance, security, and cost management of AI model access.","avoid_when":"You are not on Azure, need a cloud-agnostic solution, or want a ready-to-deploy product rather than educational labs.","alternatives":["LiteLLM","Portkey AI Gateway","Helicone","Martian"],"af_score":64.5,"security_score":75.0,"reliability_score":null,"package_type":"mcp_server","discovery_source":["github","crates_io"],"priority":"low","status":"evaluated","version_evaluated":"unknown","last_evaluated":"2026-03-01T09:50:05.198528+00:00","performance":{"latency_p50_ms":null,"latency_p99_ms":null,"uptime_sla_percent":null,"rate_limits":null,"data_source":"llm_estimated","measured_on":null}}