Encoding MCP

Encoding MCP

The Bitmovin Encoding MCP gives your AI client read only access to your Bitmovin encodings. Ask natural-language questions like "list my encodings from the past week", "why did encoding X fail?", or "which encodings hit our region-fallback path yesterday?" — and get back grounded answers composed from our encoding APIs and event logs.

Endpoint

https://agentic.bitmovin.com/encoding/mcp

Transport: HTTP (streamable HTTP / SSE)
Auth: x-api-key header with your Bitmovin API key

The agent only sees encodings that the provided API key has access to. Use a scoped, least-privilege API key where possible.

Tools exposed

ToolDescription
ask_bitmovin_encodingAnswers questions about your encodings — list with filters, inspect detail, explain failures, summarize recent activity.

The agent can combine output from the Encoding REST API with event-log data, which is especially useful for diagnosing live encoding issues (VM provisioning delays, region fallbacks, RTMP connection events, etc.).

Setup

Claude Code

claude mcp add bitmovin-encoding \
  --transport http https://agentic.bitmovin.com/encoding/mcp \
  --header "x-api-key: YOUR_BITMOVIN_API_KEY"

Verify with:

claude mcp list

Claude Desktop

Edit ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows) and add:

{
  "mcpServers": {
    "bitmovin-encoding": {
      "url": "https://agentic.bitmovin.com/encoding/mcp",
      "headers": {
        "x-api-key": "YOUR_BITMOVIN_API_KEY"
      }
    }
  }
}

Then fully quit and relaunch Claude Desktop.

Cursor / Windsurf / other clients

Add an MCP server entry with the URL and x-api-key header. Most clients accept the same JSON shape as Claude Desktop.

ChatGPT

  1. Open chatgpt.com and go to Settings → Connectors.
  2. Click Add connector and select MCP.
  3. Enter the URL: https://agentic.bitmovin.com/encoding/mcp
  4. Under Headers, add:
    • Name: x-api-key
    • Value: your Bitmovin API key
  5. Click Save.

The ask_bitmovin_encoding tool will now appear in any new conversation.

Example prompts

  • "List my five most recent live encodings and flag any that didn't reach RUNNING."
  • "Why did encoding 4c314e8d-54b6-4b89-8cd5-204abc5c6df5 take so long to start?"
  • "Summarize encoding activity in the last 24 hours — total count, status breakdown, and any errors."
  • "Which of my encodings from yesterday fell back to an alternative cloud region?"
  • "Give me the error message and the last 10 event-log entries for encoding X."

Getting a Bitmovin API key

If you don't already have one, create an API key from the Bitmovin Dashboard under Account → API Keys. Keep the key secret — it can read and write resources scoped to the organization it belongs to.

Notes

  • The MCP server does not include a model; your MCP client's model reasons over the tool responses.
  • Calls count against your normal Bitmovin API rate limits.
  • The server is read-oriented — it does not create, modify, or delete encodings.
  • All tool invocations are traced via Langfuse so we can detect regressions and improve quality over time. Your API key and tool arguments are not retained in those traces beyond standard operational logging.