Error Reference

​

Error Response Format

{
  "error": "Human-readable error message"
}

​

Prediction-Envelope Errors

​

HTTP Status Codes

​

Common Errors

​

400: Bad Request

{
  "error": "slug parameter is required"
}

• Missing required parameters
• Invalid parameter types or values
• Malformed JSON body
• Input validation failure

​

401: Unauthorized

{
  "error": "Invalid or missing API key"
}

• Authorization: Bearer YOUR_API_KEY header not included
• API key is invalid or expired

​

402: Payment Required

{
  "status": 402,
  "error": "payment required",
  "details": "the estimated cost of your in-flight executions exceeds your current balance ($1.00). please wait for an active execution to finish or top up your balance"
}

• The requested model costs more than your current balance
• The combined estimated cost of your in-flight predictions would exceed your balance — each running prediction reserves its estimated cost until it settles

• Top up your balance, or wait for an in-flight prediction to finish. A settled prediction releases the difference between its estimate and its real cost.

​

403: Forbidden

{
  "error": "Workflow is locked and cannot be modified"
}

• Attempting to modify a locked workflow or version
• Insufficient permissions for the requested operation

​

404: Not Found

{
  "error": "Failed to fetch model: model not found"
}

• Invalid model slug
• Non-existent prediction ID, workflow ID, or execution ID
• Resource was deleted

​

429: Too Many Requests

{
  "status": 429,
  "error": "too many requests",
  "details": "Account rate limited: accounts with a balance of $10.00 or less are limited to 10 concurrent executions. Keep your balance above $10.00 for higher concurrency, or wait for an active execution to finish."
}

• Keep your balance above $10.00 — the cap does not apply above that threshold
• Otherwise, wait for an in-flight prediction to reach a terminal status and resubmit; the slot frees when it settles
• Retry with exponential backoff when details reports that too many executions were started at once — that variant is transient

​

500: Internal Server Error

{
  "error": "Failed to fetch models: internal error"
}

• Retry with exponential backoff
• If persistent, contact support

​

Product-Specific Errors

​

each::api: Prediction Errors

​

each::workflows: Execution Errors

​

each::sense: Agent Errors

​

Retry Strategy

import time
import requests

def request_with_retry(url, headers, json=None, max_retries=3):
    for attempt in range(max_retries):
        response = requests.post(url, headers=headers, json=json)
        if response.status_code in (429, 500):
            wait = 2 ** attempt
            time.sleep(wait)
            continue
        return response
    return response

async function requestWithRetry(url, options, maxRetries = 3) {
  for (let attempt = 0; attempt < maxRetries; attempt++) {
    const response = await fetch(url, options);
    if ([429, 500].includes(response.status)) {
      await new Promise((r) => setTimeout(r, 2 ** attempt * 1000));
      continue;
    }
    return response;
  }
  return response;
}