Pipelines

Updated 8 November 2026

Several steps behind one endpoint: listen to a voice note, pull out the order, check for a complaint, and speak a reply. Each step reads the request ({{input.audio}}) or an earlier step ({{steps.order.customer_name}}).

curl https://wekoya-api.wekoya.tech/v1/workflows \
  -H "Authorization: Bearer $WEKOYA_API_KEY" -H "Content-Type: application/json" \
  -d '{
    "name": "voice-orders",
    "steps": [
      { "id": "hear", "type": "transcribe", "audio": "{{input.audio}}", "language": "multi" },
      { "id": "order", "type": "chat", "model": "cfg:orders", "prompt": "{{steps.hear}}", "json": true },
      { "id": "complaint", "type": "decide", "state": "{{steps.hear}}",
        "questions": { "complaint": { "type": "noul", "instructions": "Is the customer complaining?" } } },
      { "id": "confirm", "type": "speak", "text": "Thank you {{steps.order.customer_name}}, your order is on its way." }
    ]
  }'

# Run it: one call, every step in order
curl https://wekoya-api.wekoya.tech/v1/workflows/voice-orders/run \
  -H "Authorization: Bearer $WEKOYA_API_KEY" -H "Content-Type: application/json" \
  -d '{ "input": { "audio": "data:audio/ogg;base64,..." } }'

Step types: transcribe, chat (any model or a cfg: name, can read a file with attach), decide (Jev), speak, image (draw from a prompt, or edit pictures given in images, such as an outfit onto {{input.photo}}) and music.
The reply lists every step with its output, time and cost. If a step fails, the pipeline stops there and status is failed, unless the step is marked optional.
Every run costs money: each step is billed like a direct request, and the reply shows each step's cost. Pictures are the dearest step (about ₦57 or more each), so a pipeline makes at most 8. Creating or changing a pipeline needs an admin; running one needs the chat scope. Up to 8 steps and 25 MB of input.