{
  "manifest_version": "1.0.0",
  "template": {
    "id": "dc159d14-7738-40ca-a8a7-e45d7fa279ad",
    "slug": "llamacpp-server",
    "name": "llama.cpp Server",
    "description": "llama.cpp server with an OpenAI-compatible API running GGUF models on CPU.",
    "url": "https://railway.com/deploy/llamacpp-server",
    "upstream": {
      "image": "ghcr.io/ggml-org/llama.cpp:server-v0.5.0"
    }
  },
  "services": [
    {
      "name": "llamacpp",
      "source": {
        "image": "ghcr.io/ggml-org/llama.cpp:server-v0.5.0"
      },
      "needs_volume": true,
      "volume_mount_path": "/models",
      "http": true
    }
  ],
  "required_inputs": [
    {
      "key": "PORT",
      "service": "llamacpp",
      "description": "Provide a value for PORT.",
      "secret": false,
      "strategy": "default",
      "default": "8080"
    },
    {
      "key": "HF_HOME",
      "service": "llamacpp",
      "description": "Provide a value for HF_HOME.",
      "secret": false,
      "strategy": "default",
      "default": "/models/hf"
    },
    {
      "key": "LLAMA_CACHE",
      "service": "llamacpp",
      "description": "Provide a value for LLAMA_CACHE.",
      "secret": false,
      "strategy": "default",
      "default": "/models"
    },
    {
      "key": "LLAMA_API_KEY",
      "service": "llamacpp",
      "description": "Provide a value for LLAMA_API_KEY.",
      "secret": true,
      "strategy": "generate",
      "generate": "random_base64_32"
    },
    {
      "key": "LLAMA_ARG_HOST",
      "service": "llamacpp",
      "description": "Provide a value for LLAMA_ARG_HOST.",
      "secret": false,
      "strategy": "default",
      "default": "::"
    },
    {
      "key": "LLAMA_ARG_PORT",
      "service": "llamacpp",
      "description": "Provide a value for LLAMA_ARG_PORT.",
      "secret": false,
      "strategy": "default",
      "default": "8080"
    },
    {
      "key": "LLAMA_ARG_ALIAS",
      "service": "llamacpp",
      "description": "Provide a value for LLAMA_ARG_ALIAS.",
      "secret": false,
      "strategy": "default",
      "default": "qwen3-0.6b"
    },
    {
      "key": "OPENAI_BASE_URL",
      "service": "llamacpp",
      "description": "Provide a value for OPENAI_BASE_URL.",
      "secret": false,
      "strategy": "railway_provided",
      "railway_source": "railway_domain"
    },
    {
      "key": "LLAMA_ARG_HF_REPO",
      "service": "llamacpp",
      "description": "Provide a value for LLAMA_ARG_HF_REPO.",
      "secret": false,
      "strategy": "default",
      "default": "ggml-org/Qwen3-0.6B-GGUF"
    },
    {
      "key": "LLAMA_ARG_THREADS",
      "service": "llamacpp",
      "description": "Provide a value for LLAMA_ARG_THREADS.",
      "secret": false,
      "strategy": "default",
      "default": "4"
    },
    {
      "key": "LLAMA_ARG_CTX_SIZE",
      "service": "llamacpp",
      "description": "Provide a value for LLAMA_ARG_CTX_SIZE.",
      "secret": false,
      "strategy": "default",
      "default": "4096"
    },
    {
      "key": "OPENAI_PRIVATE_BASE_URL",
      "service": "llamacpp",
      "description": "Provide a value for OPENAI_PRIVATE_BASE_URL.",
      "secret": false,
      "strategy": "railway_provided",
      "railway_source": "railway_private_domain"
    }
  ],
  "deploy": {
    "mcp": {
      "server": "railway",
      "tool": "deploy_template",
      "args": {
        "template_code": "llamacpp-server"
      }
    },
    "cli": "railway deploy --template llamacpp-server",
    "api": {
      "method": "POST",
      "path": "/graphql/v2",
      "body": {
        "query": "mutation templateDeploy($input: TemplateDeployV2Input!) { templateDeployV2(input: $input) { projectId workflowId } }",
        "variables": {
          "input": {
            "templateId": "dc159d14-7738-40ca-a8a7-e45d7fa279ad",
            "serializedConfig": {
              "buckets": {},
              "services": {
                "50b818a9-62fb-44c3-ad46-e179e5e74649": {
                  "icon": "https://github.com/ggml-org.png",
                  "name": "llamacpp",
                  "deploy": {
                    "startCommand": null,
                    "healthcheckPath": "/health",
                    "restartPolicyType": "ON_FAILURE",
                    "restartPolicyMaxRetries": 10
                  },
                  "source": {
                    "image": "ghcr.io/ggml-org/llama.cpp:server-v0.5.0"
                  },
                  "variables": {
                    "PORT": {
                      "isOptional": false,
                      "defaultValue": "8080"
                    },
                    "HF_HOME": {
                      "isOptional": false,
                      "defaultValue": "/models/hf"
                    },
                    "LLAMA_CACHE": {
                      "isOptional": false,
                      "defaultValue": "/models"
                    },
                    "LLAMA_API_KEY": {
                      "isOptional": false,
                      "defaultValue": "{{LLAMA_API_KEY}}"
                    },
                    "LLAMA_ARG_HOST": {
                      "isOptional": false,
                      "defaultValue": "::"
                    },
                    "LLAMA_ARG_PORT": {
                      "isOptional": false,
                      "defaultValue": "8080"
                    },
                    "LLAMA_ARG_ALIAS": {
                      "isOptional": false,
                      "defaultValue": "qwen3-0.6b"
                    },
                    "OPENAI_BASE_URL": {
                      "isOptional": false,
                      "defaultValue": "https://${{RAILWAY_PUBLIC_DOMAIN}}/v1"
                    },
                    "LLAMA_ARG_HF_REPO": {
                      "isOptional": false,
                      "defaultValue": "ggml-org/Qwen3-0.6B-GGUF"
                    },
                    "LLAMA_ARG_THREADS": {
                      "isOptional": false,
                      "defaultValue": "4"
                    },
                    "LLAMA_ARG_CTX_SIZE": {
                      "isOptional": false,
                      "defaultValue": "4096"
                    },
                    "OPENAI_PRIVATE_BASE_URL": {
                      "isOptional": false,
                      "defaultValue": "http://${{RAILWAY_PRIVATE_DOMAIN}}:8080/v1"
                    }
                  },
                  "networking": {
                    "serviceDomains": {
                      "<hasDomain>:8080": {
                        "port": 8080
                      }
                    }
                  },
                  "volumeMounts": {
                    "50b818a9-62fb-44c3-ad46-e179e5e74649": {
                      "mountPath": "/models"
                    }
                  }
                }
              }
            }
          }
        }
      }
    }
  },
  "post_deploy": {
    "healthcheck": {
      "service": "llamacpp",
      "method": "GET",
      "path": "/health",
      "expect_status": 200
    }
  },
  "resources": {
    "expected_services": 1,
    "needs_volume": true
  },
  "generated_at": "2026-09-24T04:14:44.625Z",
  "generator_version": "0.1.0",
  "status": "validated",
  "validated_at": "2026-09-24T01:17:34.107Z",
  "success_rate_30d": 1,
  "validation": {
    "last_run_id": "run_d33f7da0c4b348078724",
    "checks": [
      {
        "name": "workflow_completed",
        "passed": true
      },
      {
        "name": "all_services_deployed",
        "passed": true
      },
      {
        "name": "healthcheck",
        "passed": true
      },
      {
        "name": "stays_up",
        "passed": true
      }
    ],
    "typical_ready_seconds": 92,
    "typical_build_seconds": 2,
    "typical_start_seconds": 20,
    "slowest_service": "llamacpp"
  }
}