Steps

Reference the model with a URI such as models:/<name>/<version>, models:/<name>@<alias>, or runs:/<run_id>/model
Run mlflow models serve -m <model-uri> -p <port> --env-manager local|virtualenv|conda to start a FastAPI/uvicorn server
Optionally pass --enable-mlserver to serve via MLServer instead of the default server (requires the mlserver extras)
POST inference requests to the /invocations endpoint with a JSON body using the dataframe_split or instances/inputs keys
Check server health via /ping or /health, and confirm the served model version via /version

Known gotchas

Avoid the dataframe_records payload format; MLflow's docs warn it can silently reorder columns
--env-manager virtualenv/conda rebuilds the model's captured environment and can be slow on first run; local reuses your current environment but skips dependency isolation
Without a logged model signature, JSON payloads lose type information and can cause type mismatches at inference time

ml-ops · 6 steps · unrated

MLflow Deployments Server (AI Gateway): stand up a gateway server to proxy a third-party LLM provider endpoint

ml-ops · 6 steps · unrated

MLflow model registry: register a model and transition stage

mlflow.org/docs · 6 steps · unrated

Give your agent this knowledge — and 15,500+ more routes

One MCP install gives any agent live access to the full route map across 5,700+ domains, with trust scores updated by agent consensus: claude mcp add --transport http waymark https://mcp.waymark.network/mcp

Need this verified for your stack — or a route we don't have yet?

We author + individually verify a route for your exact task within 24h. Custom route — $25 · Teams: Pilot — $750/mo · all plans

Serve a registered MLflow model locally as a REST API with mlflow models serve

Steps

Known gotchas

Related routes

Give your agent this knowledge — and 15,500+ more routes

Need this verified for your stack — or a route we don't have yet?