{"auth":{"endpoints_requiring_auth":["/admin/api/data"],"password_source":"data/admin_pass on server, or ADMIN_PASS env var","realm":"Pod Efficiency admin","type":"http_basic","username":"admin"},"description":"AI-native Lightbits Inference/Inferra Tool for Efficiency","docs_search_base":"https://docs.lightbits.ai/md/","examples":{"auto_optimize":{"existing_hardware":true,"goal":"profit","gpus_per_server":8,"num_servers":4,"preset":"l40s","workload_type":"openrouter"},"cluster_reconfigure":{"cluster_state":{"current_model":"llama8b-128k","gpu_type":"L40S","gpus":8}},"docs_search":{"k":5,"query":"install Inferra on Kubernetes"},"gpu_analyze":{"hardware":{"gpu_count":8,"gpu_memory_gb":48,"gpu_name":"L40S"}}},"gpu_analyzer":{"download_url":"https://open.lightbits.ai/sbin/gpu-analyzer.py","pop_format":"ZIP archive: manifest.json + system.json + config.json + traffic.json + provenance.json + metrics.json + attestation/digest.txt","upload_endpoints":["/api/v1/stats","/api/v1/pop"]},"mcp":{"docs":"https://github.com/LightBitsLabs/AILITE/blob/main/docs/MCP-SERVER.md","http_endpoint":"https://open.lightbits.ai/mcp","stdio_command":"python3 mcp-server/server.py --transport stdio","transport":["stdio","http"]},"name":"AILITE","rest_api":{"base_url":"https://open.lightbits.ai/api/v1","endpoints":[{"auth":"none","description":"Auto-optimize with arbitrary constraints","method":"POST","path":"/api/v1/mcp/optimize"},{"auth":"none","description":"Analyze hardware or parse .pop file","method":"POST","path":"/api/v1/mcp/gpu-analyze"},{"auth":"none","description":"Live OpenRouter demand + pricing","method":"GET","path":"/api/v1/mcp/market-intel"},{"auth":"none","description":"Search docs.lightbits.ai content","method":"GET","path":"/api/v1/mcp/docs/search"},{"auth":"none","description":"Optimal reconfiguration for managed clusters","method":"POST","path":"/api/v1/mcp/cluster-reconfigure"},{"auth":"none","description":"List MCP tools + JSON schemas","method":"GET","path":"/api/v1/mcp/tools"},{"auth":"none","description":"Health check","method":"GET","path":"/api/v1/health"},{"auth":"none","description":"Market data","method":"GET","path":"/api/v1/market"},{"auth":"none","description":"OpenRouter demand rankings","method":"GET","path":"/api/v1/demand"},{"auth":"none","description":"Infracost live pricing","method":"GET","path":"/api/v1/pricing"},{"auth":"none","description":"Benchmark data","method":"GET","path":"/api/v1/benchmarks"},{"auth":"none","description":"Instance catalog","method":"GET","path":"/api/v1/instances"}]},"service":"ailite","tools":[{"auth_required":false,"description":"Find the optimal inference-serving configuration under supplied constraints","inputs":["workload_type","existing_hardware","preset","gpu_type","gpus_per_server","num_servers","capex_usd","power_w","opex_usd_month","model","wquant","batch","max_seq_len","tp","pp","goal","user_count","activity_level","price_per_m"],"name":"auto_optimize"},{"auth_required":false,"description":"Analyze GPU hardware specs or parse a .pop file","inputs":["hardware","pop_file","pop_bytes"],"name":"gpu_analyze"},{"auth_required":false,"description":"Live OpenRouter demand, pricing trends, HuggingFace adoption, .pop leaderboard","inputs":["profiles"],"name":"market_intel"},{"auth_required":false,"description":"Search docs.lightbits.ai markdown content (BM25, section-aware)","inputs":["query","k","doc"],"name":"docs_search"},{"auth_required":false,"description":"Recommend optimal model + config for an automated inference cluster agent","inputs":["cluster_state"],"name":"cluster_reconfigure"}],"url":"https://open.lightbits.ai","version":"0.7.0"}