Getting Started with ModKie API

Welcome to ModKie. This guide covers what you need before integrating ModKie APIs into production: available models, pricing, authentication, the async request flow, rate limits, data retention, and support.

ModKie is a Kie-compatible unified AI API platform for video, image, audio, and language models — plus open-source model deployment and fine-tuning. Read this page carefully before going live.

1. Available Models & Playground

Browse the latest supported models on the Market:

👉 https://modkie.com/market

• We continuously onboard new models as they become stable.
• Each model links to a Playground at /playground/{slug} where you can test parameters and output formats before calling the API.
• The Playground is the fastest way to understand model behavior and copy a working createTask payload.

See the full curated catalog in Market Models.

2. Pricing

The up-to-date per-model credit rates are on the pricing page:

👉 https://modkie.com/pricing

• Credits are charged only when a generation task completes successfully.
• Failed tasks are not charged.
• Pricing may change as upstream providers adjust costs — always refer to the pricing page for current numbers.

3. Creating and Securing Your API Key

Create and manage API keys here:

👉 https://modkie.com/settings/apikeys

Security notes:

• Never expose your API key in frontend code (browsers, mobile apps, public repositories).
• Treat your API key as a secret.

Per-key protections available in key settings:

• Rate limits — optional hourly, daily, and total usage caps
• IP whitelist — restrict access to approved server IPs

These help prevent accidental overuse and unauthorized requests.

4. Required Request Headers

Every API request must include:

Authorization: Bearer <YOUR_API_KEY>
Content-Type: application/json

If headers are missing or incorrect, you may receive:

{ "code": 401, "msg": "You do not have access permissions" }

Always verify headers first when debugging authentication issues.

5. Logs & Task Details

Inspect all historical API tasks here:

👉 https://modkie.com/logs

For each task you can view:

• Creation time
• Model used
• Input parameters
• Task status
• Credit consumption
• Final results or error details

If you suspect incorrect credit usage, the logs page is the source of truth.

6. Data Retention Policy

Download and store results on your side if you need long-term access.

7. Asynchronous Task Model

All generation tasks on ModKie are asynchronous.

A successful createTask request returns:

• HTTP 200
• A taskId

:::warning A 200 OK response only means the task was queued. It does not mean generation finished. :::

To get the final result, you must either:

• Pass top-level callBackUrl on createTask (see webhooks), or
• Poll recordInfo using the taskId

Quick reference:

POST https://modkie.com/api/v1/jobs/createTask
GET  https://modkie.com/api/v1/jobs/recordInfo?taskId={taskId}

8. Rate Limits & Concurrency

Default platform limits:

• Up to 20 new generation requests per 10 seconds per account
• Excess requests return HTTP 429 and are not queued
• Per-key hourly/daily caps can be configured in API key settings

If you consistently hit 429 errors, contact support to request a higher limit.

9. Developer Support

Recommended channels:

• Discord — community and technical help
• Telegram — community updates
• Email: [email protected] (slower than chat channels)

10. Stability Expectations

ModKie provides access to top-tier AI APIs at competitive pricing through upstream relays.

That said:

• We are not perfect
• Overall stability may be slightly lower than calling official providers directly
• This is a conscious trade-off for cost and unified integration

In practice, ModKie is stable enough for production workloads, but we prefer setting realistic expectations upfront.

11. About ModKie

ModKie is 「开源模型的 Kie.ai + 模型落地顾问」 — a unified AI API platform that also helps teams deploy, fine-tune, and integrate open-source models.

Next steps:

1. Market Quickstart — first API call in minutes
2. API Platform overview — endpoints and billing model
3. createTask — queue a generation job