vllm-sr 0.2.0.dev20260318063308

vLLM Semantic Router - Intelligent routing for Mixture-of-Models

vLLM Semantic Router

Intelligent Router for Mixture-of-Models (MoM).

GitHub: https://github.com/vllm-project/semantic-router

Quick Start

Installation

# Install from PyPI
pip install vllm-sr

# Or install from source (development)
cd src/vllm-sr
pip install -e .

Usage

# Start the router (includes dashboard and first-run setup)
HF_TOKEN=hf_xxx vllm-sr serve

# Start an isolated second local stack on offset host ports
VLLM_SR_STACK_NAME=lane-b VLLM_SR_PORT_OFFSET=200 HF_TOKEN=hf_xxx vllm-sr serve

# Open the dashboard
# http://localhost:8700
# second stack example: http://localhost:8900

# Optional: open the dashboard in your browser
vllm-sr dashboard

# View logs
vllm-sr logs router
vllm-sr logs envoy
vllm-sr logs dashboard

# Check status
vllm-sr status

# Stop
vllm-sr stop

If you start in an empty directory, vllm-sr serve bootstraps a minimal workspace and opens the dashboard in setup mode. Configure your first model there, then activate routing.

Local dashboard state is persisted under .vllm-sr/dashboard-data/ and bind-mounted into the container at /app/data. User accounts, evaluation history, and ML pipeline artifacts survive vllm-sr stop followed by a new vllm-sr serve as long as that workspace directory is kept.

To run parallel local stacks from the same machine or multiple worktrees, set VLLM_SR_STACK_NAME and VLLM_SR_PORT_OFFSET before vllm-sr serve, vllm-sr status, vllm-sr dashboard, and vllm-sr stop. The stack name isolates container and network names, and the port offset shifts the published host ports while keeping internal container ports unchanged.

Advanced YAML-first setup

# Validate a hand-authored canonical config before serving
vllm-sr validate config.yaml

vllm-sr init was removed in v0.3. Author config.yaml directly using the canonical version/listeners/providers/routing/global layout, or migrate an older file with vllm-sr config migrate --config old-config.yaml. Router-wide defaults come from the router itself and can be overridden under global:.

Features

• Router: Intelligent request routing based on intent classification
• Envoy Proxy: High-performance proxy with ext_proc integration
• Dashboard: Web UI for monitoring and testing (http://localhost:8700)
• Metrics: Prometheus metrics endpoint (http://localhost:9190/metrics)

Endpoints

After running vllm-sr serve, the following endpoints are available:

Endpoint	Port	Description
Dashboard	8700	Web UI for monitoring and Playground
API	8888*	Chat completions API (configurable in config.yaml)
Metrics	9190	Prometheus metrics
gRPC	50051	Router gRPC (internal)
Jaeger UI	16686	Distributed tracing UI
Grafana (embedded)	8700	Dashboards at /embedded/grafana
Prometheus UI	9090	Metrics storage and querying

*Default port, configurable via listeners in config.yaml

Observability

vllm-sr serve automatically starts the observability stack:

• Jaeger: Distributed tracing embedded at http://localhost:8700/embedded/jaeger (also available directly at http://localhost:16686)
• Grafana: Pre-configured dashboards embedded at http://localhost:8700/embedded/grafana
• Prometheus: Metrics collection at http://localhost:9090

Note: Grafana is optimized for embedded access through the dashboard. For the best experience, use http://localhost:8700/embedded/grafana where anonymous authentication is pre-configured.

Tracing is enabled by default. Traces are visible in Jaeger under the vllm-sr service name.

Configuration

Plugin Configuration

The CLI supports configuring plugins in your routing decisions. Plugins are per-decision behaviors that customize request handling (security, caching, customization, debugging).

Supported Plugin Types:

• semantic-cache - Cache similar requests for performance
• jailbreak - Detect and block adversarial prompts
• pii - Detect and enforce PII policies
• system_prompt - Inject custom system prompts
• header_mutation - Add/modify HTTP headers
• hallucination - Detect hallucinations in responses
• router_replay - Record routing decisions for debugging

Plugin Examples:

Each example shows the plugin list inside a canonical routing.decisions[] entry.

1. semantic-cache - Cache similar requests:

routing:
  decisions:
    - name: "cached-route"
      plugins:
        - type: "semantic-cache"
          configuration:
            enabled: true
            similarity_threshold: 0.92  # 0.0-1.0, higher = more strict
            ttl_seconds: 3600  # Optional: cache TTL in seconds

2. jailbreak - Block adversarial prompts:

routing:
  decisions:
    - name: "guarded-route"
      plugins:
        - type: "jailbreak"
          configuration:
            enabled: true
            threshold: 0.8  # Optional: detection sensitivity 0.0-1.0

3. pii - Enforce PII policies:

routing:
  decisions:
    - name: "pii-route"
      plugins:
        - type: "pii"
          configuration:
            enabled: true
            threshold: 0.7  # Optional: detection sensitivity 0.0-1.0
            pii_types_allowed: ["EMAIL_ADDRESS"]  # Optional: list of allowed PII types

4. system_prompt - Inject custom instructions:

routing:
  decisions:
    - name: "persona-route"
      plugins:
        - type: "system_prompt"
          configuration:
            enabled: true
            system_prompt: "You are a helpful assistant."
            mode: "replace"  # "replace" (default) or "insert" (prepend)

5. header_mutation - Modify HTTP headers:

routing:
  decisions:
    - name: "header-route"
      plugins:
        - type: "header_mutation"
          configuration:
            add:
              - name: "X-Custom-Header"
                value: "custom-value"
            update:
              - name: "User-Agent"
                value: "SemanticRouter/1.0"
            delete:
              - "X-Old-Header"

6. hallucination - Detect hallucinations:

routing:
  decisions:
    - name: "fact-check-route"
      plugins:
        - type: "hallucination"
          configuration:
            enabled: true
            use_nli: false  # Optional: use NLI for detailed analysis
            hallucination_action: "header"  # "header", "body", or "none"

7. router_replay - Record decisions for debugging:

routing:
  decisions:
    - name: "debug-route"
      plugins:
        - type: "router_replay"
          configuration:
            enabled: true
            max_records: 200  # Optional: max records in memory (default: 200)
            capture_request_body: false  # Optional: capture request payloads (default: false)
            capture_response_body: false  # Optional: capture response payloads (default: false)
            max_body_bytes: 4096  # Optional: max bytes to capture (default: 4096)

Router replay records are exposed through:

• GET /v1/router_replay?limit=20&offset=0&search=req-123&decision=foo&model=bar&cache_status=cached - List recent records with pagination metadata. Default page size is 20; larger limit values are capped at 100.
• GET /v1/router_replay/aggregate?search=req-123&decision=foo&model=bar&cache_status=cached - Return summary and chart aggregates for the filtered replay set.
• GET /v1/router_replay/{id} - Fetch a single replay record.

If a replay page would exceed the ext-proc gRPC message budget, the router returns 413 Payload Too Large instead of failing the stream.

Validation Rules:

• Plugin Type: Must be one of: semantic-cache, jailbreak, pii, system_prompt, header_mutation, hallucination, router_replay
• enabled: Must be a boolean (required for most plugins)
• threshold/similarity_threshold: Must be a float between 0.0 and 1.0
• max_records/max_body_bytes: Must be a positive integer
• ttl_seconds: Must be a non-negative integer
• pii_types_allowed: Must be a list of strings (if provided)
• system_prompt: Must be a string (if provided)
• mode: Must be "replace" or "insert" (if provided)

CLI Commands:

# Validate configuration (including plugins)
vllm-sr validate

# Migrate older configs to the canonical contract
vllm-sr config migrate --config old-config.yaml

File Descriptor Limits

The CLI automatically sets file descriptor limits to 65,536 for Envoy proxy. To customize:

export VLLM_SR_NOFILE_LIMIT=100000  # Optional (min: 8192)
vllm-sr serve

License

Apache 2.0