Quick Start

Tokaify API is a unified AI gateway and usage management platform. It supports OpenAI-compatible calls, multi-model access, token-based permissions, balance management, and request-level auditing. The flow below follows a typical MaaS quickstart, adapted to how Tokaify actually works.

1. Sign In to the Platform

After signing in to the Tokaify console, open the Dashboard and confirm the following:

• Whether your current balance is sufficient
• Whether the API Address can be copied directly
• Whether today's usage, RPM, and TPM look normal
• Whether platform announcements, maintenance notices, or quick links require attention

If this is your first integration, review both the dashboard and the wallet first so you can confirm account status before debugging.

2. Review Available Models

Open the Model Gallery to review the models, pricing, and capability tags available to your account.

Recommended evaluation order:

1. Filter by task type first, such as text, vision, multimodal, image, or video.
2. Then compare pricing, context length, and capability tags.
3. Finally, copy the standard model identifier for the model parameter in your code.

If you plan to use Tokaify in production, avoid typing model names manually. Copy them directly from the model details or API response.

3. Test the Model in Playground

In the Playground, you can validate model behavior before writing any code.

• Choose a group and model to confirm your account can access it
• Enter prompts or multi-turn messages to test output quality
• Adjust parameters such as temperature to match the generation style you want
• Inspect the preview request, actual request, and response content to verify the payload format

4. Create an API Key

Go to Token Management and click Add Token to create your credential.

At minimum, configure these fields:

• Token name: use a project- or environment-based name such as web-prod or agent-test
• Group: select the appropriate business group
• Model restrictions: allow only the models your workload actually needs
• Quota / expiration / IP restrictions: add them based on your risk level

After creating the token, copy the full secret. Use the entire sk-... value during integration and do not remove the prefix manually.

5. Make Your First API Call

5.1 Verify the model list first

curl https://api.tokaify.com/v1/models \
  -H "Authorization: Bearer sk-your-token"

If this returns a valid model list, your Base URL and API Key are both working.

5.2 Call a chat model over REST

curl https://api.tokaify.com/v1/chat/completions \
  -H "Authorization: Bearer sk-your-token" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "your-model-id",
    "messages": [
      {
        "role": "system",
        "content": "You are a helpful assistant."
      },
      {
        "role": "user",
        "content": "Describe the Tokaify API in one sentence."
      }
    ]
  }'

5.3 Call the API with the OpenAI SDK

First install the official Python SDK:

pip install --upgrade openai

Then point base_url to Tokaify's OpenAI-compatible endpoint:

from openai import OpenAI

client = OpenAI(
    api_key="sk-your-token",
    base_url="https://api.tokaify.com/v1",
)

response = client.chat.completions.create(
    model="your-model-id",
    messages=[
        {
            "role": "user",
            "content": "Describe the Tokaify API in one sentence."
        }
    ],
)

print(response.choices[0].message.content)

If you are integrating through clients such as Codex, Cursor, VS Code, or Claude Code instead of your own application code, jump to Scenario Examples and follow the field mappings for the relevant tool.

6. Review Results, Usage, and Balance

After the first successful request, return to the console and verify the full loop:

• In Usage Log, confirm the request succeeded, which model was actually used, and whether billing looks correct
• In Task Log, confirm async jobs completed for image, video, or other task-style APIs
• In Wallet, confirm balance and spending updates
• In Account Center, maintain your basic account settings

Common Entry Points

Next Steps