The Small Language Model API allows you to integrate Ollama-based language models into your project for free. I created it because I wanted to be able to write webapps that use these models without requiring users to have lots of available processing power. Please use it kindly as I am hosting this service on my own hardware.

Synchronous Completion

• HTTPS Only (insecure requests return a 521 error)
• The request will take ~30-60s for the LLM to process
• This method does not support streaming

curl https://slm.robog.net/respond \
  --request POST \
  --header "Content-Type: application/json" \
  --data @- << EOF
    [
      {
        "role": "user",
        "content": "Tell me a joke!"
      }
    ]
EOF

Streamed Completion

// TODO