AGIOne User Guide for End Users

This guide is written for first-time AGIOne users. It walks you through the full basic workflow: signing in, claiming free quota, trying a model in the web playground, and calling the model with curl. Usernames, passwords, and API keys are shown as placeholders only.

Applicable Roles

• End User trying and calling a model
• Model Provider reviewing the resulting customer call records

1. Before You Start

Prepare the following information before using AGIOne.

Item	Example or Description
Platform URL	https://agione.pro/
Username	{USERNAME}
Password	{PASSWORD}
Model API Key	{API_KEY}, copied from the Quick Start page

This guide uses Qwen3.5-27b as an example. Model availability and provider-instance identifiers can vary by environment.

Item	Value
Model Name	Qwen3.5-27b
Model Identifier	{MODEL_IDENTIFIER}
Protocol	openai/chat_completions
API Endpoint	https://agione.pro/hyperone/xapi/api/v1/chat/completions

Provider-instance call identifiers can vary by environment, provider, or region. Copy the identifier shown in the current environment from the selected provider card or Quick Start before calling.

2. Sign In to AGIOne

2.1 Open the Sign-In Page

Enter the following address in your browser:

https://agione.pro/user/login

You will see the AGIOne sign-in page.

2.2 Fill In the Sign-In Form

Fill in the form in the following order.

Step	Action	Value
1	Enter your username in Username or email	{USERNAME}
2	Enter your password in Password	{PASSWORD}
3	Select the agreement checkbox	Check it
4	Click Sign in	Wait for the platform to open

If a privacy policy or service terms dialog appears, click Agree.

2.3 Confirm That Sign-In Succeeded

After signing in successfully, you should see a user avatar or initial in the upper-right corner. The left-side menu should show entries such as Discover, Playground, Studio, Usage & Earnings, and My Calls.

3. Open the Model List

3.1 Open Model Services

After signing in, open:

https://agione.pro/modelone/store/model

You can also navigate through the page menu:

Model Services > Discover > Models

3.2 Find Qwen3.5-27b

Search for Qwen3.5-27b in the model list. The list is sorted by Newest, so use the search box instead of relying on the model's position in the current list. After you find it, click View on the model row.

4. Claim Free Quota

The free quota entry is on the provider card in the model detail page, next to the Playground button.

4.1 Open the Model Detail Page

After opening the Qwen3.5-27b detail page, confirm that you can see the following information.

Check Item	Expected Value
Model Name	Qwen3.5-27b
Model ID	qwen/qwen3.5-27b
Provider Card Call Identifier	{MODEL_IDENTIFIER}
Quota Button	Claim Free Quota
Trial Entry	Playground

4.2 Click to Claim

On the provider card, click:

Claim Free Quota

After the quota is claimed successfully, the page shows a success message and the button changes to:

Claimed

4.3 Confirm That the Quota Was Claimed

Any of the following signs means the quota has been claimed successfully.

Success Sign	Description
The page shows Successfully claimed	The platform has completed the claim
The button changes to Claimed	The current account has already claimed the free quota for this model
The model detail page still shows Claimed after reopening it	The claim status has been saved

5. Try the Model in the Web Playground

The web playground is the easiest way to try a model. You do not need to write code.

5.1 Open Playground from the Detail Page

On the Qwen3.5-27b model detail page, click the following button on the provider card:

Playground

You can also open it from the left-side menu:

Playground > Text

If you open it from the left-side menu, select the target model manually. If you open it from the model detail page, the current model is usually preselected.

5.2 Send a Test Message

After entering the text playground page, follow these steps.

Step	Action	Recommended Value
1	Confirm Model Sub-Type	LLM
2	Confirm or select the model	The provider model for Qwen3.5-27b
3	Confirm Protocol	openai/chat_completions
4	Adjust parameters if needed	Beginners can keep the default values
5	Enter a question in the input box at the bottom	Introduce AGIOne in one sentence.
6	Click the send button	Wait for the model response

If the generated response appears in the chat area, the playground call has succeeded.

5.3 Select the Playground That Matches the Model Modality

The model's capabilities and input/output modalities determine which playground to use. Do not validate image, video, or audio models with text-model parameters.

Modality	Menu Path	Validate
Text / chat	Playground > Text	Prompt, Temperature, Top-P, Max Tokens, response, and latency
Image / multimodal	Playground > Images	Authorized image, prompt, size, count, output image, and request ID
Video	Playground > Video	Authorized short video, frame sampling, duration, result, and asynchronous latency
Speech / audio	Playground > Audio	Audio format, sample rate, language, voice, and text or audio output

For an image model, choose an available image or multimodal model, upload sanitized and authorized material, set the prompt and size, and send the request.

For a video model, first use a short sanitized sample to validate frame sampling, output format, and asynchronous results before increasing input size.

For an audio model, confirm input format, language, and output type. Do not upload original recordings that contain customer privacy.

6. Call the Model with curl

Use curl when you want to call the model from a command line, script, or backend service.

6.1 Open the Quick Start Page

Go back to the model detail page and click:

Quick Start

On the Quick Start page, you can find the following information.

Information	Purpose
Model ID Call Identifier	The provider-specific value used in the model field
Base URL	The API service base address
Path	The protocol path appended to the base URL
Full URL	The complete curl request URL
Model API Key	The authentication key selected in the AUTHENTICATION section
Curl Example	A ready-to-use reference command

The API key in the screenshot is redacted. In actual use, copy your own API key from the page.

6.2 Copy the API Key

Find Model API Keys in the AUTHENTICATION section, select an authorized key, then click Copy.

Please note:

Note	Description
Do not share your API key with others	The API key can be used to call models under your account
Do not commit it to public documents or code repositories	Store it as an environment variable when possible
This guide uses {API_KEY}	Replace it with your real key when calling the API

6.3 curl Example

Replace {API_KEY} and {MODEL_IDENTIFIER} with values copied from the current environment, then run the command in your terminal.

curl -X POST "https://agione.pro/hyperone/xapi/api/v1/chat/completions" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer {API_KEY}" \
  -d '{
    "stream": true,
    "model": "{MODEL_IDENTIFIER}",
    "messages": [
      {
        "role": "user",
        "content": "Hello"
      }
    ]
  }'

6.4 Confirm That the API Call Succeeded

When the call succeeds, your terminal returns generated model output. Common signs of success include:

Check Item	Success Sign
Response content	The response includes generated text
Model field	The response includes the current environment's model identifier
No authentication error	No Unauthorized or Invalid API key error appears
No quota error	No insufficient quota message appears

6.5 Quick Practice

Go to AGIOne Best Practices.

7. View Your Call Records

To confirm whether the model was actually called, check the call logs.

7.1 Open Call Logs

Open the following menu from the left side:

My Calls > Call Logs

7.2 Filter Call Records

Step	Action
1	Select the call time range
2	Enter the model name or model ID
3	Select the call status, such as success or failure
4	Click Search
5	Check the call time, model, status, token usage, and latency in the list
6	To view more information, click Details on the target row

8. FAQ

8.1 What Should I Do If Sign-In Fails?

Possible Cause	What to Do
Incorrect username or password	Confirm {USERNAME} and {PASSWORD}
Agreement checkbox not selected	Select the checkbox and sign in again
Agreement dialog not confirmed	Click Agree

8.2 What Should I Do If I Cannot Find Qwen3.5-27b?

What to Do	Description
Search for Qwen3.5-27b in the model list	Use the search box to locate it quickly
Make sure you are viewing All Models	Avoid filtering to private models only
Change sorting or page through the list	The model may not be visible in the first screen

8.3 What Should I Do If the Claim Button Cannot Be Clicked?

Page Status	Description
The button shows Claimed	The current account has already claimed it
Claim Free Quota is missing	The model may not support free quota, or your account may not have permission
Nothing happens after clicking	Refresh the page and try again, or confirm that you are still signed in

8.4 What Should I Do If the curl Call Fails?

Error Symptom	What to Check
Authentication failure	Check whether the API key is complete and whether the header is Authorization: Bearer {API_KEY}
Model not found	Check whether model matches the provider card call identifier copied from the current environment
Incorrect request URL	Check whether the URL is https://agione.pro/hyperone/xapi/api/v1/chat/completions
Invalid JSON	Check quotation marks, commas, and braces
Insufficient quota	Confirm that free quota has been claimed, or check your account quota

9. Quick Reference

Feature	Entry
Sign in	https://agione.pro/user/login
Model list	https://agione.pro/modelone/store/model
Qwen3.5-27b detail page	Model Services > Discover > Models > Qwen3.5-27b > View
Claim free quota	Qwen3.5-27b detail page provider card > Claim Free Quota
Web playground	Qwen3.5-27b detail page provider card > Playground
API instructions	Qwen3.5-27b detail page > Quick Start
Call logs	My Calls > Call Logs

Completion Checklist

Purpose: These are the exit criteria for the current feature task. Use them to decide whether the result is observable and reviewable and whether you can continue to the next step in the scenario. They do not repeat the procedure; if any item fails, follow the troubleshooting section below.

Check	Pass Criteria
1	Sign-in succeeds and the intended model can be found.
2	Free quota or other authorized quota is available.
3	A playground request and an API request both return valid responses.
4	The API call appears in Call Logs with the expected model and status.

Troubleshooting

Symptom	Check First
The intended model cannot be found	Model publication state, public/private scope, account authorization, and marketplace filters
Playground works but the API request fails	Personal Key, endpoint, request headers, protocol, model identifier, and rate limit
The request succeeds but no call log appears	Account and project scope, time range, model filter, and processing delay

User Manual

• Models
• My Calls Overview