API calls

See Authentication for access methods and OpenAPI endpoints for paths, parameters, and response fields.

Request format

The base URL is https://api.chaintable.com, with paths starting with /v1/. Send GET parameters in the query string. For POST requests, send a JSON object with Content-Type: application/json. For example, list tables in the demo Space anonymously. Replace demo with the public Space ID you want to query:

curl --get 'https://api.chaintable.com/v1/table/list' \
  --data-urlencode 'space=demo' \
  --data-urlencode 'limit=10'

On success, the table list is in data.items.

Example request

For example, invoke demo.usdc_amount saved in Create a Function with the input 1500000:

curl 'https://api.chaintable.com/v1/function/invoke' \
  -H 'Content-Type: application/json' \
  --data '{"id": "demo.usdc_amount", "arguments": [1500000]}'

id is the full Function ID. arguments is an array of values in parameter order. Pass [] even when the Function takes no arguments. A successful response contains the Function’s return value in data.result:

{
  "code": 0,
  "message": "ok",
  "data": {
    "success": true,
    "result": 1.5,
    "error": null
  }
}

Handle responses

Responses contain code, message, and data. HTTP 200 with code: 0 means the API request succeeded. If a request fails, use the HTTP status, numeric error code, and message to diagnose the cause.

HTTP status	Action
`400`	Check request parameters and use `message` to correct the request
`401`	Check that the required credentials are provided and valid
`403`	Check resource permissions, permissions on the target account, and required quota
`404`	Check the resource ID
`409`	Check for a duplicate resource or whether its current state allows the operation
`429`	The request was rate-limited; reduce the request rate and wait before retrying
`5xx`	Server or upstream failure; keep the error details and `X-Request-ID` response header for troubleshooting

Compute Units

API Key calls consume the owning account’s Compute Units (CU) based on request processing time: each millisecond costs 1 CU, with a minimum of 100 CU per call.

CU consumed = max(100, round(request processing time in milliseconds))

After credential, allowance, and rate limit checks pass, responses with HTTP status below 500 are billed, including 4xx. Requests that fail these checks or return 5xx are not billed. View your CU allowance, usage, and reset time under Capacity → Openapi. See Plans and usage.

Overview

Quickstart

Guides

Reference

Request format

Example request

Handle responses

Compute Units