AI Market Gateway Usage and Integration Guide - AI Market
AIAI Market Docs

AI Market Gateway Usage and Integration Guide

Create a token after registration to start using models.

、Integration。The sk-your-key ,。
Last updated: 2026-05-03 14:54

AI Market Gateway Usage and Integration Guide

Updated:2026-05-03

Docs AI Market Gateway,Integration GPT、Claude、Gemini 3.1、Image generation。The sk-your-key ,Console。

Claude Code Quick Start

Claude Code AI Market https://ai.ai-market.com,Do not add /v1

npm install -g @anthropic-ai/claude-code
$env:ANTHROPIC_BASE_URL="https://ai.ai-market.com"
$env:ANTHROPIC_AUTH_TOKEN="YOUR_API_KEY"
claude

macOS / Linux export set the same variables.

Codex Quick Start

Codex CLI OpenAI https://ai.ai-market.com/v1,Recommended models gpt-5.3-codex

npm install -g @openai/codex@latest
codex --version

~/.codex/config.toml(Windows %USERPROFILE%\.codex\config.toml):

model_provider = "nbapi"
model = "gpt-5.3-codex"
[model_providers.nbapi]
name = "AI Market"
base_url = "https://ai.ai-market.com/v1"
wire_api = "responses"
requires_openai_auth = true

~/.codex/auth.json {"OPENAI_API_KEY":"YOUR_API_KEY"} codex 。 Key。

1. Quick Integration Info

/Console:

https://ai.ai-market.com/

OpenAI Endpoint:

https://ai.ai-market.com/v1

Claude / Anthropic Endpoint:

https://ai.ai-market.com

Common headers:

Authorization: Bearer sk-your-key
Content-Type: application/json

Anthropic / Claude Messages can also use:

x-api-key: sk-your-key
anthropic-version: 2023-06-01
Content-Type: application/json

30-Second Preflight Check

  1. Prefer OpenAI Compatible / OpenAI API / Custom OpenAI URL
  2. GPT、Gemini、Image generationall use https://ai.ai-market.com/v1
  3. Claude Code Anthropic Messages native mode uses https://ai.ai-market.com,Do not append /v1
  4. ID Copy, gpt-5.5gpt-5.3-codexgemini-3.1-pro-previewclaude-opus-4-7
  5. Streaming Output, stream: falsePONG before running long tasks。
  6. Never send the full key when troubleshooting, 4 。

2. Create Token

  1. Sign in https://ai.ai-market.com/
  2. Console/API Key Page。
  3. Create a token and select its group.
  4. Copy sk-... ,Put it in your client, code, or environment variables.
  5. /v1/models ,Integration。

、、、Docs。。

3. 、Usage Guide

This section helps new users get started.

3.1 Registration and Sign-in

  1. Open the website:https://ai.ai-market.com
  2. Click Register and follow the instructions.
  3. Sign inConsole。
  4. Sign in、。

3.4 Usage Order

New users should follow this order:

Register -> Sign inConsole -> Create Token ->  -> Integration -> 

Common page uses:

PagePurpose
HomeView site information and available models
ConsoleView account, usage, and tokens
ModelsView supported models
/API Key sk- API key starting with; address:/console/token
/Review call logs and usage

If setup fails, send support these three items:

Software name
Base URL
Model name

Never send the full API key. Redact sk- the middle of the token。

4. Group Selection

Calls usually fail because the token group does not match the model.

GPT / OpenAI :

ChatGpt 1
ChatGpt Standard
default

Gemini 3.1 :

Gemini
Gemini 3.1
default

Claude :

Claude code
ClaudeCall
default
CC3

Create Token:GPT GPT ,Gemini Gemini ,Claude Code Claude code 。、。

GPT Text and Code Models

gpt-5.5
gpt-5.4
gpt-5.4-mini
gpt-5.3-codex
gpt-5.2

Recommended choices:

ScenarioRecommended models
General capability, long text, and complex tasksgpt-5.5
High-quality general Q&Agpt-5.4
Low cost and speedgpt-5.4-mini
Programming, code review, and repository analysisgpt-5.3-codex
Reliable fallbackgpt-5.2

GPT Image generation

gpt-image-2

POST /v1/images/generations

Gemini 3.1

gemini-3.1-pro-preview
gemini-3.1-pro-high
gemini-3.1-pro-low
gemini-3.1-flash-lite

Recommended choices:

ScenarioRecommended models
Gemini 3.1 gemini-3.1-pro-preview
/gemini-3.1-pro-high
Gemini 3.1 Callgemini-3.1-pro-low
gemini-3.1-flash-lite

Claude

Recommended Claude 4.x :

claude-opus-4-7
claude-opus-4-7-max
claude-opus-4-7-xhigh
claude-opus-4-7-high
claude-opus-4-7-medium
claude-opus-4-7-low
claude-opus-4-7-thinking
claude-sonnet-4-6
claude-opus-4-6
claude-opus-4-5-20251101
claude-haiku-4-5-20251001

Recommended choices:

ScenarioRecommended models
Claude Code、claude-opus-4-7 / claude-opus-4-7-high
Taskclaude-opus-4-7-max / claude-opus-4-7-xhigh
Daily Claude useclaude-sonnet-4-6
Lightweight and low costclaude-haiku-4-5-20251001

GPT、Claude OpenAI Models。Recommended models;Call /v1/models

6. Check Available Models

curl https://ai.ai-market.com/v1/models \
  -H "Authorization: Bearer sk-your-key"

If the target model is missing,the token group, authorization, quota, or model availability may be the cause.

7. OpenAI Compatible Chat API

For most clients choose OpenAI CompatibleOpenAI APICustom OpenAI URL and enter the following values:

Base URL: https://ai.ai-market.com/v1
API Key: sk-your-key
Model: gpt-5.5

cURL

curl https://ai.ai-market.com/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer sk-your-key" \
  -d '{
    "model": "gpt-5.5",
    "messages": [
      {"role": "system", "content": "You are a helpful assistant."},
      {"role": "user", "content": "Reply PONG"}
    ],
    "stream": false
  }'

Python

Install:

pip install openai

Call:

from openai import OpenAI

client = OpenAI(
    api_key="sk-your-key",
    base_url="https://ai.ai-market.com/v1",
)

response = client.chat.completions.create(
    model="gpt-5.5",
    messages=[
        {"role": "system", "content": "You are a helpful assistant."},
        {"role": "user", "content": "Reply PONG"},
    ],
)

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

Node.js

Install:

npm install openai

Call:

import OpenAI from "openai";

const openai = new OpenAI({
  apiKey: "sk-your-key",
  baseURL: "https://ai.ai-market.com/v1",
});

const response = await openai.chat.completions.create({
  model: "gpt-5.5",
  messages: [
    { role: "system", content: "You are a helpful assistant." },
    { role: "user", content: "Reply PONG" },
  ],
});

console.log(response.choices[0].message.content);

Streaming Output

import OpenAI from "openai";

const openai = new OpenAI({
  apiKey: "sk-your-key",
  baseURL: "https://ai.ai-market.com/v1",
});

const stream = await openai.chat.completions.create({
  model: "gpt-5.4",
  messages: [{ role: "user", content: "Write a 100-word product description" }],
  stream: true,
});

for await (const chunk of stream) {
  process.stdout.write(chunk.choices?.[0]?.delta?.content || "");
}

Codex CLI Integration AI Market

Codex Suitable for、、 Bug、Scenario。AI Market OpenAI Integration Codex , gpt-5.3-codex;, gpt-5.5gpt-5.4 gpt-5.2

Prepare:

Base URL: https://ai.ai-market.com/v1
API Key: sk-your-key
Model: gpt-5.3-codex
: ChatGpt 1 / ChatGpt Standard / default

Install Codex CLI:

npm install -g @openai/codex

macOS Homebrew:

brew install codex

。Windows PowerShell :

$env:OPENAI_API_KEY="sk-your-key"

Windows PowerShell :

[Environment]::SetEnvironmentVariable("OPENAI_API_KEY", "sk-your-key", "User")

macOS / Linux:

export OPENAI_API_KEY="sk-your-key"

Codex :

Windows: %USERPROFILE%\.codex\config.toml
macOS/Linux: ~/.codex/config.toml

: OpenAI 。

model = "gpt-5.3-codex"
openai_base_url = "https://ai.ai-market.com/v1"

: AI Market Provider, OpenAI Gateway。

model = "gpt-5.3-codex"
model_provider = "cr-ai"

[model_providers.cr-ai]
name = "AI Market"
base_url = "https://ai.ai-market.com/v1"
env_key = "OPENAI_API_KEY"

Start test:

codex

Contents,Task:

Reply only PONG,。

, Codex AI Market GatewayCall。 Codex Task,:

,、Contents,。

Common issues:

SymptomAction
Sign in OPENAI_API_KEYconfig.toml ; Codex。
401 / Unauthorized、、Copy;ConsoleCreate Token。
404 / ,; /v1/models
429 / Reduce concurrency, retry later, or use another group.
Long task interrupted gpt-5.3-codex gpt-5.5,。

Codex App IDE API Key Base URL,; ChatGPT Sign in、 Base URL,Integration AI Market。

OpenClaw()Integration AI Market

OpenClaw()/ AI ,Integration Telegram、Discord、WhatsApp、Console。AI Market OpenAI GatewayIntegration OpenClaw , AI Market provider, cr-ai

Install OpenClaw

OpenClaw Node.js,Node 24 ,Node 22.14+ 。Install:

macOS / Linux:

curl -fsSL https://openclaw.ai/install.sh | bash

Windows PowerShell:

iwr -useb https://openclaw.ai/install.ps1 | iex

openclaw onboard --install-daemon

Gateway:

openclaw gateway status
openclaw dashboard

AI Market Provider

AI Market :

API Key: sk-your-key
Base URL: https://ai.ai-market.com/v1
Provider ID: cr-ai

Key ,。

macOS / Linux:

export CRAI_API_KEY="sk-your-key"

Windows PowerShell:

$env:CRAI_API_KEY="sk-your-key"

OpenClaw provider。:

{
  agents: {
    defaults: {
      model: {
        primary: "cr-ai/gpt-5.5",
        fallbacks: ["cr-ai/gpt-5.4", "cr-ai/gpt-5.2"]
      },
      imageGenerationModel: {
        primary: "cr-ai/gpt-image-2"
      },
      models: {
        "cr-ai/gpt-5.5": { alias: "AI Market GPT-5.5" },
        "cr-ai/gpt-5.4": { alias: "AI Market GPT-5.4" },
        "cr-ai/gpt-5.4-mini": { alias: "AI Market Mini" },
        "cr-ai/gpt-5.3-codex": { alias: "AI Market Codex" },
        "cr-ai/gpt-5.2": { alias: "AI Market GPT-5.2" },
        "cr-ai/gpt-image-2": { alias: "AI Market Image" }
      }
    }
  },
  models: {
    mode: "merge",
    providers: {
      "cr-ai": {
        baseUrl: "https://ai.ai-market.com/v1",
        apiKey: "${CRAI_API_KEY}",
        api: "openai-completions",
        models: [
          { id: "gpt-5.5", name: "GPT-5.5" },
          { id: "gpt-5.4", name: "GPT-5.4" },
          { id: "gpt-5.4-mini", name: "GPT-5.4 Mini" },
          { id: "gpt-5.3-codex", name: "GPT-5.3 Codex" },
          { id: "gpt-5.2", name: "GPT-5.2" },
          { id: "gpt-image-2", name: "GPT Image 2", input: ["text", "image"] }
        ]
      }
    }
  }
}

Refresh models after setup:

openclaw models list --provider cr-ai
openclaw models set cr-ai/gpt-5.5
openclaw models status
openclaw gateway restart

OpenClaw :

/model list
/model cr-ai/gpt-5.5
/model cr-ai/gpt-5.3-codex
/model status

Recommended models:

ScenarioOpenClaw Model reference
Daily chat, writing, and translationcr-ai/gpt-5.5 / cr-ai/gpt-5.4
Fast low-cost repliescr-ai/gpt-5.4-mini
、、Agent Taskcr-ai/gpt-5.3-codex
cr-ai/gpt-5.2
Image generationcr-ai/gpt-image-2

Note:AI Market OpenClaw Integration /v1

https://ai.ai-market.com/v1

OpenClaw Troubleshooting

SymptomAction
cr-ai/* models.providers.cr-aiopenclaw models list --provider cr-ai
Authentication missing CRAI_API_KEY whether the environment variable is active, apiKey ${CRAI_API_KEY}
401 / Unauthorized、Copy; AI Market Console sk-
404 / ,; /v1/models
Model is not allowedagents.defaults.models ,; cr-ai/
Base URL /v1https://ai.ai-market.com/v1
Image generation imageGenerationModel.primary cr-ai/gpt-image-2,。

Hermes Agent Integration AI Market

Hermes Agent / AI Agent , Custom endpoint Integration OpenAI 。AICodeMirror Sign in https://www.aicodemirror.com/dashboard/hermes, Hermes ; AI Market ,:

Provider: Custom endpoint
API base URL: https://ai.ai-market.com/v1
API Key: sk-your-key
Model name: gpt-5.5

Install Hermes

Linux / macOS / WSL2:

curl -fsSL https://raw.githubusercontent.com/NousResearch/hermes-agent/main/scripts/install.sh | bash
source ~/.bashrc
hermes --version

Windows Install,Install WSL2, WSL2 Install。

Method One:Integration

hermes model

Choose as prompted Custom endpoint / Self-hosted,then enter:

API base URL: https://ai.ai-market.com/v1
API key: sk-your-key
Model name: gpt-5.5

Test after setup:

hermes chat -q " AI Market Integration Hermes"

hermes chat -m gpt-5.3-codex -q "Inspect the project structure and suggest improvements"

Method Two:

~/.hermes/.env,:

CRAI_API_KEY=sk-your-key

~/.hermes/config.yaml

model:
  provider: custom
  base_url: https://ai.ai-market.com/v1
  api_key: "${CRAI_API_KEY}"
  default: gpt-5.5

hermes config set model.provider custom
hermes config set model.base_url https://ai.ai-market.com/v1
hermes config set model.api_key sk-your-key
hermes config set model.default gpt-5.5
hermes config check

Gateway, AI Market :

custom_providers:
  - name: cr-ai
    base_url: https://ai.ai-market.com/v1
    key_env: CRAI_API_KEY
    api_mode: chat_completions
    models:
      gpt-5.5:
        context_length: 128000
      gpt-5.4:
        context_length: 128000
      gpt-5.4-mini:
        context_length: 128000
      gpt-5.3-codex:
        context_length: 128000
      gpt-5.2:
        context_length: 128000
      gemini-3.1-pro-preview:
        context_length: 128000
      claude-opus-4-7:
        context_length: 200000

Provider:

/model custom:cr-ai:gpt-5.5
/model custom:cr-ai:gpt-5.3-codex
/model custom:cr-ai:gemini-3.1-pro-preview

Recommended models:

ScenarioHermes
、、gpt-5.5 / gpt-5.4
Fast low-cost repliesgpt-5.4-mini
Agent、、gpt-5.3-codex
Text、gemini-3.1-pro-preview
Claude Taskclaude-opus-4-7 / claude-sonnet-4-6

Note:gpt-image-2 Image generation, Hermes ;Image generation。

Hermes Troubleshooting

SymptomAction
OpenRouter hermes config set model.provider customhermes model Custom endpoint
401 / UnauthorizedAI Market 、, sk-
404 / Copy, gpt-5.3-codex,。
Connection or path errorBase URL https://ai.ai-market.com/v1,Do not omit /v1
/model AI Market/model Provider;, hermes model custom_providers
Windows Install WSL2 Hermes, Linux Install。

8. Gemini 3.1 Integration

Gemini 3.1 Integration OpenAI 。 Google , OpenAI :

Base URL: https://ai.ai-market.com/v1
API Key: sk-your-key
Model: gemini-3.1-pro-preview

cURL :

curl https://ai.ai-market.com/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer sk-your-key" \
  -d '{
    "model": "gemini-3.1-pro-preview",
    "messages": [
      {"role": "user", "content": "Describe Gemini 3.1 in three sentences"}
    ]
  }'

Python :

from openai import OpenAI

client = OpenAI(
    api_key="sk-your-key",
    base_url="https://ai.ai-market.com/v1",
)

response = client.chat.completions.create(
    model="gemini-3.1-flash-lite",
    messages=[{"role": "user", "content": "Reply PONG"}],
)

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

Gemini API / Google AI Studio , Google 。currently recommended OpenAI Compatible Integration Gemini 3.1。

9. Claude IntegrationMethod One:OpenAI

OpenAI , Claude:

Base URL: https://ai.ai-market.com/v1
API Key: sk-your-key
Model: claude-opus-4-7

cURL :

curl https://ai.ai-market.com/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer sk-your-key" \
  -d '{
    "model": "claude-opus-4-7",
    "messages": [
      {"role": "user", "content": "Reply PONG in Chinese"}
    ]
  }'

Suitable for Cherry Studio、Chatbox、LobeChat、Open WebUI、Cline、Roo Code、Continue、NextChat OpenAI 。

10. Claude IntegrationMethod Two:Anthropic Messages

Suitable for Anthropic 。

cURL :

curl https://ai.ai-market.com/v1/messages \
  -H "x-api-key: sk-your-key" \
  -H "anthropic-version: 2023-06-01" \
  -H "content-type: application/json" \
  -d '{
    "model": "claude-opus-4-7",
    "max_tokens": 1024,
    "messages": [
      {"role": "user", "content": "Reply PONG"}
    ]
  }'

Python Anthropic SDK :

pip install anthropic
from anthropic import Anthropic

client = Anthropic(
    api_key="sk-your-key",
    base_url="https://ai.ai-market.com",
)

message = client.messages.create(
    model="claude-opus-4-7",
    max_tokens=1024,
    messages=[{"role": "user", "content": "Reply PONG"}],
)

print(message.content[0].text)

11. Claude Code Integration

Claude Code ANTHROPIC_BASE_URL /v1,:

https://ai.ai-market.com

macOS / Linux

export ANTHROPIC_BASE_URL="https://ai.ai-market.com"
export ANTHROPIC_API_KEY="sk-your-key"
export ANTHROPIC_DEFAULT_OPUS_MODEL="claude-opus-4-7"
export ANTHROPIC_DEFAULT_SONNET_MODEL="claude-sonnet-4-6"
export ANTHROPIC_DEFAULT_HAIKU_MODEL="claude-haiku-4-5-20251001"
claude

Windows PowerShell

$env:ANTHROPIC_BASE_URL="https://ai.ai-market.com"
$env:ANTHROPIC_API_KEY="sk-your-key"
$env:ANTHROPIC_DEFAULT_OPUS_MODEL="claude-opus-4-7"
$env:ANTHROPIC_DEFAULT_SONNET_MODEL="claude-sonnet-4-6"
$env:ANTHROPIC_DEFAULT_HAIKU_MODEL="claude-haiku-4-5-20251001"
claude

/ Authorization: Bearer,:

export ANTHROPIC_AUTH_TOKEN="sk-your-key"

Key,。

Claude Code

~/.claude/settings.json

{
  "env": {
    "ANTHROPIC_BASE_URL": "https://ai.ai-market.com",
    "ANTHROPIC_API_KEY": "sk-your-key",
    "ANTHROPIC_DEFAULT_OPUS_MODEL": "claude-opus-4-7",
    "ANTHROPIC_DEFAULT_SONNET_MODEL": "claude-sonnet-4-6",
    "ANTHROPIC_DEFAULT_HAIKU_MODEL": "claude-haiku-4-5-20251001"
  }
}

Claude Code :

/status
/model

、。

12. GPT Image 2 Image generation

cURL :

curl https://ai.ai-market.com/v1/images/generations \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer sk-your-key" \
  -d '{
    "model": "gpt-image-2",
    "prompt": "A clean premium business portrait poster, white background, realistic photography",
    "n": 1,
    "size": "1024x1024",
    "response_format": "b64_json"
  }'

Python :

from openai import OpenAI
import base64

client = OpenAI(
    api_key="sk-your-key",
    base_url="https://ai.ai-market.com/v1",
)

image = client.images.generate(
    model="gpt-image-2",
    prompt="A clean premium business portrait poster, white background, realistic photography",
    size="1024x1024",
    n=1,
)

item = image.data[0]
if getattr(item, "b64_json", None):
    with open("output.png", "wb") as f:
        f.write(base64.b64decode(item.b64_json))
else:
    print(item.url)

Note:gpt-image-2background: transparent

13. /

Vision GPT、Claude、Gemini 。OpenAI :

curl https://ai.ai-market.com/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer sk-your-key" \
  -d '{
    "model": "gpt-5.4",
    "messages": [
      {
        "role": "user",
        "content": [
          {"type": "text", "text": "Describe this image"},
          {
            "type": "image_url",
            "image_url": {
              "url": "https://example.com/image.jpg"
            }
          }
        ]
      }
    ]
  }'

Base64 Data URL, image_url.url

14. Responses API Integration

OpenAI Responses API。,:

curl https://ai.ai-market.com/v1/responses \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer sk-your-key" \
  -d '{
    "model": "gpt-5.5",
    "input": "Reply PONG"
  }'

Responses API , /v1/chat/completions,Call。

15. Common Software Settings

Cherry Studio / Chatbox / NextChat / LobeChat

Provider: OpenAI Compatible
API Host / Base URL: https://ai.ai-market.com/v1
API Key: sk-your-key
Model: gpt-5.5 / gpt-5.4 / gemini-3.1-pro-preview / claude-opus-4-7

Open WebUI

OPENAI_API_BASE_URL=https://ai.ai-market.com/v1
OPENAI_API_KEY=sk-your-key

gpt-5.5
gpt-5.4
gpt-5.3-codex
gemini-3.1-pro-preview
claude-opus-4-7

Cline / Roo Code / Continue

Provider: OpenAI Compatible
Base URL: https://ai.ai-market.com/v1
API Key: sk-your-key
Model ID: claude-opus-4-7  gpt-5.3-codex

Anthropic ,Claude can also use:

Anthropic Base URL: https://ai.ai-market.com
Anthropic API Key: sk-your-key
Model: claude-opus-4-7

Cursor / Windsurf and IDEs

Custom OpenAI URL:

OpenAI Compatible Base URL: https://ai.ai-market.com/v1
API Key: sk-your-key
Model: gpt-5.3-codex

gpt-5.3-codexclaude-opus-4-7claude-sonnet-4-6

16. Practical Skills:Reliable Calls and High-Quality Output

These Skill Copy, Cherry Studio、Chatbox、Claude Code、Cline、Roo Code、Continue 。The approach follows OpenAI、Anthropic、Google :verify the token, then define the task, input, constraints, output format, and acceptance criteria。

Skill 1:Three-Step Token Check

Suitable forIntegration,“”。

curl https://ai.ai-market.com/v1/models \
  -H "Authorization: Bearer sk-your-key"
  1. TextCall:
curl https://ai.ai-market.com/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer sk-your-key" \
  -d '{
    "model": "gpt-5.4-mini",
    "messages": [
      {"role": "user", "content": "Reply only PONG"}
    ],
    "stream": false
  }'
  1. 。、,、、。

Skill 2:Choose Models by Task

ScenarioRecommended models
、、gpt-5.4-mini / Gemini Flash
gpt-5.5 / gpt-5.4 / gemini-3.1-pro-preview,Suitable forTask
、Agent、IDE gpt-5.3-codex / claude-opus-4-7 / claude-sonnet-4-6
Image generation、gpt-image-2、、、
Vision GPT / Claude / Gemini URL Base64 Data URL

Skill 3:General Prompt Template

:[, /  / ]

:
[]

:
[、、、]

:
1. Do not invent information。
2. Mark uncertain items as Needs confirmation。
3. Be concise and prioritize actionable conclusions。

:
1. Give the conclusion first。
2. Then give steps or a table。
3. End with risks and next steps。

:
Check that no key constraints were missed。

Skill 4:JSON / JSON and Table Extraction

Suitable for、、、。

Extract structured data from the text below. Output valid JSON only, without Markdown or explanation.

Fields:
- name: string
- email: string,use null when missing
- amount: number,use null when missing
- status: one of ["paid", "unpaid", "unknown"]
- risk: one of ["normal", "suspicious", "need_check"]

Text:
[]

response_format, JSON :

{
  "response_format": {
    "type": "json_object"
  }
}

JSON Schema ,,。

Skill 5:Code and Agent Workflow

Suitable for Claude Code、Cline、Roo Code、Continue、Cursor、Windsurf。

Complete the task using this process:
1. Read the relevant files and configuration and describe the current state。
2. Give a brief plan listing files to change and why。
3. Keep the change minimal and avoid unrelated files。
4. Run available tests, lint, or minimal verification。
5. Report changes, verification, and remaining risks。

Task:
[/]

Claude Code “ -> -> -> ”。, /clear ,。

Skill 6:gpt-image-2 Commercial Image Prompt

a high-resolution realistic commercial image。

:[ /  / Scenario]
Purpose:[portrait poster / product image / promotional graphic / infographic]
:[front-facing half-body / centered / overhead / space for text]
:[white / office / technological but clean]
:[soft even / studio / natural shadows]
:[realistic photography / premium business / clean and simple]
:[if text is needed, specify English copy, hierarchy, color, and position]
:[No transparent background, watermark, extra text, or distortion]
Aspect ratio:[1:1 / 16:9 / 9:16]

,;。“、Aspect ratio”,。

Skill 7:Gemini 3.1 Long Documents and Multimodal

,。

Task:
[ /  /  /  / ]

:
[Text、Docs]

:
1.  3 。
2. 。
3. ,。
4. 。

Task。“”,second round“”,third round“”,。

Skill 8:Cost, Speed, and Reliability

  1. gpt-5.4-mini Flash ,。
  2. Troubleshooting stream: false,Streaming Output。
  3. ,、、;。
  4. max_tokens,、、JSON 。
  5. 502 upstream_errorgpt-5.5 gpt-5.4, Gemini Pro Flash。
  6. API Key 、; 4 。

Skill 9:Support Troubleshooting Script

 4 :
1. Base URL,Note Key。
2.  ID。
3. 。
4. :Cherry Studio / Chatbox / Claude Code / Cline / Cursor / 。

Task, /v1/models  PONG  Key、、、。

17. Troubleshooting

Symptom
401 unauthorizedKey 、、Copy Key;OpenAI Authorization: Bearer;Anthropic x-api-key ANTHROPIC_API_KEY
403 / , GPT ChatGpt Standard,Gemini Gemini
model not found /v1/models ,Copy ID
insufficient quota
502 upstream_errorgpt-5.5 gpt-5.4
Claude Code 404ANTHROPIC_BASE_URL /v1 https://ai.ai-market.com
Gemini Google Gemini OpenAI Compatible,Base URL /v1
gpt-image-2
Streaming Output stream: false Call

18. Support Quick Replies

Endpoint:https://ai.ai-market.com/v1
:OpenAI Compatible
Key: sk- 
:gpt-5.5 / gpt-5.4 / gpt-5.3-codex / gemini-3.1-pro-preview / claude-opus-4-7

Claude Code :

ANTHROPIC_BASE_URL=https://ai.ai-market.com
ANTHROPIC_API_KEY= sk- 
Recommended models:claude-opus-4-7  claude-sonnet-4-6

Image generation:

Endpoint:https://ai.ai-market.com/v1/images/generations
:gpt-image-2
:1024x1024

19. Security Recommendations

  1. Create Token。
  2. ,。
  3. Docs、、 Key。
  4. GPT、Claude、Gemini,。

20. Official References

21. Contact

、CallTroubleshooting,。、、、,。

QQ

:1058444275。Scan to join the AI Market support group.

AI Market QQ

AI Market 。,。

AI Market

Security reminder:Do not share full API keys or passwords. Provide only redacted keys, request time, and error screenshots.