sk-your-key ,。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"
claudemacOS / 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/v1Claude / Anthropic Endpoint:
https://ai.ai-market.comCommon headers:
Authorization: Bearer sk-your-key
Content-Type: application/jsonAnthropic / Claude Messages can also use:
x-api-key: sk-your-key
anthropic-version: 2023-06-01
Content-Type: application/json30-Second Preflight Check
- Prefer
OpenAI Compatible/OpenAI API/Custom OpenAI URL。 - GPT、Gemini、Image generationall use
https://ai.ai-market.com/v1。 - Claude Code Anthropic Messages native mode uses
https://ai.ai-market.com,Do not append/v1。 - ID Copy,
gpt-5.5、gpt-5.3-codex、gemini-3.1-pro-preview、claude-opus-4-7。 - Streaming Output,
stream: false,PONGbefore running long tasks。 - Never send the full key when troubleshooting, 4 。
2. Create Token
- Sign in
https://ai.ai-market.com/。 - Console/API Key Page。
- Create a token and select its group.
- Copy
sk-...,Put it in your client, code, or environment variables. -
/v1/models,Integration。
、、、Docs。。
3. 、Usage Guide
This section helps new users get started.
3.1 Registration and Sign-in
- Open the website:
https://ai.ai-market.com - Click Register and follow the instructions.
- Sign inConsole。
- Sign in、。
3.4 Usage Order
New users should follow this order:
Register -> Sign inConsole -> Create Token -> -> Integration -> Common page uses:
| Page | Purpose |
|---|---|
| Home | View site information and available models |
| Console | View account, usage, and tokens |
| Models | View 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 nameNever 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
defaultGemini 3.1 :
Gemini
Gemini 3.1
defaultClaude :
Claude code
ClaudeCall
default
CC3Create Token:GPT GPT ,Gemini Gemini ,Claude Code Claude code 。、。
5. Recommended Models
GPT Text and Code Models
gpt-5.5
gpt-5.4
gpt-5.4-mini
gpt-5.3-codex
gpt-5.2Recommended choices:
| Scenario | Recommended models |
|---|---|
| General capability, long text, and complex tasks | gpt-5.5 |
| High-quality general Q&A | gpt-5.4 |
| Low cost and speed | gpt-5.4-mini |
| Programming, code review, and repository analysis | gpt-5.3-codex |
| Reliable fallback | gpt-5.2 |
GPT Image generation
gpt-image-2:
POST /v1/images/generationsGemini 3.1
gemini-3.1-pro-preview
gemini-3.1-pro-high
gemini-3.1-pro-low
gemini-3.1-flash-liteRecommended choices:
| Scenario | Recommended models |
|---|---|
| Gemini 3.1 | gemini-3.1-pro-preview |
| / | gemini-3.1-pro-high |
| Gemini 3.1 Call | gemini-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-4-5-20251001Recommended choices:
| Scenario | Recommended models |
|---|---|
| Claude Code、 | claude-opus-4-7 / claude-opus-4-7-high |
| Task | claude-opus-4-7-max / claude-opus-4-7-xhigh |
| Daily Claude use | claude-sonnet-4-6 |
| Lightweight and low cost | claude-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 Compatible、OpenAI API、Custom OpenAI URL and enter the following values:
Base URL: https://ai.ai-market.com/v1
API Key: sk-your-key
Model: gpt-5.5cURL
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 openaiCall:
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 openaiCall:
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-5.5、gpt-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 / defaultInstall Codex CLI:
npm install -g @openai/codexmacOS 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:
codexContents,Task:
Reply only PONG,。, Codex AI Market GatewayCall。 Codex Task,:
,、Contents,。Common issues:
| Symptom | Action |
|---|---|
| Sign in | OPENAI_API_KEY, config.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 | bashWindows PowerShell:
iwr -useb https://openclaw.ai/install.ps1 | iex:
openclaw onboard --install-daemonGateway:
openclaw gateway status
openclaw dashboardAI Market Provider
AI Market :
API Key: sk-your-key
Base URL: https://ai.ai-market.com/v1
Provider ID: cr-aiKey ,。
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 restartOpenClaw :
/model list
/model cr-ai/gpt-5.5
/model cr-ai/gpt-5.3-codex
/model statusRecommended models:
| Scenario | OpenClaw Model reference |
|---|---|
| Daily chat, writing, and translation | cr-ai/gpt-5.5 / cr-ai/gpt-5.4 |
| Fast low-cost replies | cr-ai/gpt-5.4-mini |
| 、、Agent Task | cr-ai/gpt-5.3-codex |
cr-ai/gpt-5.2 | |
| Image generation | cr-ai/gpt-image-2 |
Note:AI Market OpenClaw Integration /v1:
https://ai.ai-market.com/v1OpenClaw Troubleshooting
| Symptom | Action |
|---|---|
cr-ai/* | models.providers.cr-ai , openclaw models list --provider cr-ai。 |
| Authentication missing | CRAI_API_KEY whether the environment variable is active, apiKey ${CRAI_API_KEY}。 |
401 / Unauthorized | 、Copy; AI Market Console sk- 。 |
404 / | ,; /v1/models 。 |
Model is not allowed | agents.defaults.models ,; cr-ai/ 。 |
Base URL /v1, https://ai.ai-market.com/v1。 | |
| Image generation | imageGenerationModel.primary cr-ai/gpt-image-2,。 |
Hermes Agent Integration AI Market
Hermes Agent / AI Agent , Custom endpoint Integration OpenAI 。AICodeMirror Sign in https://www.aicodemirror.com/dashboard/hermes, Hermes ; AI Market ,:
Provider: Custom endpoint
API base URL: https://ai.ai-market.com/v1
API Key: sk-your-key
Model name: gpt-5.5Install Hermes
Linux / macOS / WSL2:
curl -fsSL https://raw.githubusercontent.com/NousResearch/hermes-agent/main/scripts/install.sh | bash
source ~/.bashrc
hermes --versionWindows Install,Install WSL2, WSL2 Install。
Method One:Integration
hermes modelChoose as prompted Custom endpoint / Self-hosted,then enter:
API base URL: https://ai.ai-market.com/v1
API key: sk-your-key
Model name: gpt-5.5Test after setup:
hermes chat -q " AI Market Integration Hermes":
hermes chat -m gpt-5.3-codex -q "Inspect the project structure and suggest improvements"Method Two:
~/.hermes/.env,:
CRAI_API_KEY=sk-your-key ~/.hermes/config.yaml:
model:
provider: custom
base_url: https://ai.ai-market.com/v1
api_key: "${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 checkGateway, 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: 200000Provider:
/model custom:cr-ai:gpt-5.5
/model custom:cr-ai:gpt-5.3-codex
/model custom:cr-ai:gemini-3.1-pro-previewRecommended models:
| Scenario | Hermes |
|---|---|
| 、、 | gpt-5.5 / gpt-5.4 |
| Fast low-cost replies | gpt-5.4-mini |
| Agent、、 | gpt-5.3-codex |
| Text、 | gemini-3.1-pro-preview |
| Claude Task | claude-opus-4-7 / claude-sonnet-4-6 |
Note:gpt-image-2 Image generation, Hermes ;Image generation。
Hermes Troubleshooting
| Symptom | Action |
|---|---|
| OpenRouter | hermes config set model.provider custom, hermes model Custom endpoint。 |
401 / Unauthorized | AI Market 、, sk- 。 |
404 / | Copy, gpt-5.3-codex,。 |
| Connection or path error | Base 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-previewcURL :
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-4-7cURL :
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 anthropicfrom 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.commacOS / 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"
claudeWindows 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-2 , background: 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-4-7Open 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-4-7Cline / 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-codexAnthropic ,Claude can also use:
Anthropic Base URL: https://ai.ai-market.com
Anthropic API Key: sk-your-key
Model: claude-opus-4-7Cursor / 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-codex、claude-opus-4-7、claude-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"- 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
}'- 。、,、、。
Skill 2:Choose Models by Task
| Scenario | Recommended 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
-
gpt-5.4-miniFlash ,。 - Troubleshooting
stream: false,Streaming Output。 - ,、、;。
-
max_tokens,、、JSON 。 -
502 upstream_error,gpt-5.5gpt-5.4, Gemini Pro Flash。 - 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 unauthorized | Key 、、 | 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_error | , gpt-5.5 gpt-5.4 | |
| Claude Code 404 | ANTHROPIC_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-4-7Claude Code :
ANTHROPIC_BASE_URL=https://ai.ai-market.com
ANTHROPIC_API_KEY= sk-
Recommended models:claude-opus-4-7 claude-sonnet-4-6Image generation:
Endpoint:https://ai.ai-market.com/v1/images/generations
:gpt-image-2
:1024x102419. Security Recommendations
- Create Token。
- ,。
- Docs、、 Key。
- 。
- GPT、Claude、Gemini,。
20. Official References
- OpenAI :https://developers.openai.com/api/docs/guides/external-models#custom-endpoints
- OpenAI Chat Completions API:https://developers.openai.com/api/reference/resources/chat
- OpenAI Image generation:https://developers.openai.com/api/docs/guides/images-vision#generate-or-edit-images
- OpenAI :https://developers.openai.com/api/docs/guides/tools-image-generation#tool-options
- OpenAI :https://developers.openai.com/api/docs/guides/prompt-engineering
- OpenAI :https://developers.openai.com/api/docs/guides/structured-outputs
- OpenAI GPT Image :https://developers.openai.com/cookbook/examples/multimodal/image-gen-models-prompting-guide
- OpenAI Codex Quick Start:https://developers.openai.com/codex/quickstart
- OpenAI Codex :https://developers.openai.com/codex/config-reference
- OpenClaw Quick Start:https://docs.openclaw.ai/start/getting-started
- OpenClaw OpenAI Provider:https://docs.openclaw.ai/providers/openai
- OpenClaw Provider:https://docs.openclaw.ai/concepts/model-providers
- OpenClaw models :https://docs.openclaw.ai/cli/models
- AICodeMirror Hermes (Sign in):https://www.aicodemirror.com/dashboard/hermes
- Hermes Agent Install:https://hermes-agent.nousresearch.com/docs/getting-started/installation
- Hermes Agent :https://hermes-agent.nousresearch.com/docs/user-guide/configuration
- Hermes Agent AI Providers:https://hermes-agent.nousresearch.com/docs/integrations/providers
- Gemini OpenAI :https://ai.google.dev/gemini-api/docs/openai
- Gemini :https://ai.google.dev/gemini-api/docs/prompting-strategies
- Gemini :https://ai.google.dev/gemini-api/docs/structured-output
- Gemini TextStreaming Output:https://ai.google.dev/gemini-api/docs/text-generation
- Anthropic Messages API :https://platform.claude.com/docs/en/build-with-claude/working-with-messages
- Anthropic Claude :https://platform.claude.com/docs/en/build-with-claude/prompt-engineering/overview
- Claude Code :https://code.claude.com/docs/en/best-practices
- Claude Code :https://code.claude.com/docs/en/env-vars
- Claude Code LLM Gateway :https://code.claude.com/docs/en/llm-gateway
- Claude Code :https://code.claude.com/docs/en/model-config
21. Contact
、CallTroubleshooting,。、、、,。
:1058444275。Scan to join the AI Market support group.
AI Market 。,。
Security reminder:Do not share full API keys or passwords. Provide only redacted keys, request time, and error screenshots.