# OneKey Gateway AI Agent Gateway to Commercial APIs, Skills, MCPs and more using just one access key for most benefits. [GitHub](https://github.com/aiagenta2z/onekey-gateway)|[Document](http://www.deepnlp.org/doc/onekey_gateway)|[AI Agent Marketplace](http://www.deepnlp.org/store/ai-agent)|[AI Agent Search](http://www.deepnlp.org/search/agent) | [AI Agent A2Z Deployment](https://deepnlp.org/doc/agent_mcp_deployment) | [Agent A2Z Payment](https://www.deepnlp.org/doc/agent_a2z_payment) **Onekey Gateway** focuses on commercial agent APIs/MCPs/Skills, which typically require multiple subscriptions or access keys for web search,image generation, financial data, 3D rendering. Our mission is to help users authenticate once and enjoy the most benefits, discounted, pay-as-you-go usage. Accessing Agent APIs, MCPs, and LLMs across 30+ categories and 100+ tools via our credit system, you can eliminate multiple expensive subscriptions and maximize your benefits. More SOTA agent/models are available in the [DeepNLP AI Agent Mareketplace](https://deepnlp.org/store/ai-agent). Top Ranked Agent API(Skills/CLI)/MCPs Dashboard Image generation ## Overview [OneKey Gateway Document](https://www.deepnlp.org/doc/onekey_gateway) OneKey Gateway provides CLI tool, Restful API, MCP, Skills and Python/Typescript methods for Agent and human to use commercial Agent APIs. | Gateway Type | Description | BASE URL | CLI | |-------------|---------------------------------------------------------------------------------------------------------------|----------------------------------------|----------------| | Agent APIs | Allow your AI Agent to Access APIs [OneKey Agent Router Doc](https://www.deepnlp.org/doc/onekey_agent_router) | https://agent.deepnlp.org/agent_router | `npx onekey agent` | | MCP | Allow your AI clients to connect to commercial Http MCP [OneKey MCP Router Doc](https://www.deepnlp.org/doc/onekey_mcp_router) | https://agent.deepnlp.org/mcp | `npx onekey mcp` | | LLM | Allow your AI clients to call LLM endpoint proxy [OneKey LLM Router Doc](https://www.deepnlp.org/doc/onekey_llm_router) | https://agent.deepnlp.org/llm | `npx onekey llm` | Want to register your API to OneKey Agent Gateway? See the [Registry Documents](docs/onekey-gateway/registry/README.md) and top ranked onekey gateway APIs [Agent API Skills MCP Calls Ranking](https://deepnlp.org/agent/rankings) | Gateway Type | Description | |---------------------------------------------------|------------------------------------------------------------------------------------------------------------------------------------------------------| | Register Your New API | OneKey Gateway Registry [Documents GitHub](docs/onekey-gateway/registry/README.md) | ### Architecture This is an example of how onekey-gateway can help you and your AI agents to use commercial API,Skills and MCPs in one single gateway. #### OneKey Gateway Agent/Skills/MCP Conversion Matrix Register and maintain your APIs/MCP server only once, AI Agent Marketplace registry, convert to other formats 10x faster, supported APIs/CLIs/Skills/Rest API/MCPs | From \ To | API | CLI | Skills | Routed API | MCPs (StreamingHTTP) | MCPs (Stdio Local) | |---------------------|-----|-----|--------|------------|----------------------|--------------------| | **API** | - | ✅ | ✅ | ✅ | ✅ | - | | **CLI** | - | - | ✅ | - | - | ✅ | | **Skills** | - | - | - | - | ✅ | ✅ | | **Routed API** | - | - | - | - | - | - | | **MCPs (Streaming HTTP)** | - | - | - | - | - | - | | **MCPs (Stdio Local)** | - | - | - | - | - | - | ```shell ┌─────────────┐ │ Client │ │ (App / CLI) │ └─────┬───────┘ │ ▼ ┌───────────────────┐ │ OneKey Gateway │ │───────────────────│ │ • Authentication │ │ • Security │ │ • Billing/Quotas │ │ • Rate Limiting │ └─────┬─────────────┘ │ ▼ ┌──────────────────┐ │ OneKey Router │ ├──────────────────┤ │ • Routes requests│ │ • Load balances │ │ • Policy checks │ └─────┬────────────┘ ┌────────────┼────────────┐ ▼ ▼ ▼ ┌────────────────┐ ┌────────────────┐ ┌────────────────┐ │ Agent Router │ │ MCP Router │ │ LLM Router │ │ (Agents / │ │ (Tools / MCPs) │ │ (LLM Providers)│ │ Skills) │ │ │ │ │ └─────┬──────────┘ └─────┬──────────┘ └─────┬──────────┘ │ │ │ ▼ ▼ ▼ ┌─────────────┐ ┌─────────────┐ ┌─────────────┐ │ Execution │ │ Execution │ │ Execution │ │ Layer │ │ Layer │ │ Layer │ │ • APIs │ │ • MCP │ │ • LLM │ │ • Orchest │ │Servers/Tools│ │ Providers │ └─────┬───────┘ └─────┬───────┘ └─────┬───────┘ ▼ ▼ ▼ ┌──────────────────────────────────────────────┐ │ Provider APIs / External Services │ │ • LLM APIs, SaaS tools, Databases, etc. │ └──────────────────────────────────────────────┘ ``` ### Install ```bash npm -g install @aiagenta2z/onekey-gateway ``` Test installation if run onekey got `permission denied`, run from npx ``` npx onekey ``` ### OneKey Access Key **Step 1**. Get AI Agent A2Z OneKey access key You can get the OneKey Gateway [Access Keys](https://deepnlp.org/workspace/keys). **Step 2**. Set OneKey Access ```shell export DEEPNLP_ONEKEY_ROUTER_ACCESS=your_access_key ``` For demo purpose, use the demo key for demo results. ```shell export DEEPNLP_ONEKEY_ROUTER_ACCESS=BETA_TEST_KEY_MAR_2026 ``` The CLI reads the access key from `DEEPNLP_ONEKEY_ROUTER_ACCESS`. If not set, it falls back to `BETA_TEST_KEY_MARCH_2026`. ## Quickstart ### OneKey Gateway CLI #### Example 1: Google Maps Your Agent will search Google Maps to find places, rankings, etc. prompt: New York City Italian Restaurants ```bash npx onekey agent google-maps/google-maps maps_search_places '{"query":"New York City Italian Restaurants"}' ``` Result ```shell {"success":true,"data":"[\"{\\n \\\"places\\\": [\\n {\\n \\\"name\\\": \\\"Carmine's - Time Square\\\",\\n \\\"formatted_address\\\": \\\"200 W 44th St, New York, NY 10036, USA\\\",\\n \\\"location\\\": {\\n \\\"lat\\\": 40.757498,\\n \\\"lng\\\": -73.986654\\n },\\n \\\"place_id\\\": \\\"ChIJR9So-lRYwokRX1xEjA0rChA\\\",\\n \\\"rating\\\": 4.5,\\n \\\"types\\\": [\\n \\\"establishment\\\",\\n \\\"food\\\",\\n \\\"point_of_interest\\\",\\n \\\"restaurant\\\"\\n ]\\n },\\n {\\n \\\"name\\\": \\\"Osteria La Baia\\\",\\n \\\"formatted_address\\\": \\\"129 W 52nd St, New York, NY 10019, USA\\\",\\n \\\"location\\\": {\\n \\\"lat\\\": 40.7618881,\\n \\\"lng\\\": -73.9809702\\n },\\n \\\"place_id\\\": \\\"ChIJo9FTvDBZwokRFAZ0i4jbLuk\\\",\\n \\\"rating\\\": 4.9,\\n \\\"types\\\": [\\n \\\"bar\\\",\\n \\\"establishment\\\",\\n \\\"food\\\",\\n \\\"point_of_interest\\\",\\n \\\"restaurant\\\"\\n ]\\n },\\n {\\n \\\"name\\\": \\\"Tony's Di Napoli\\\",\\n \\\"formatted_address\\\": \\\"147 W 43rd St, New York, NY 10036, USA\\\",\\n \\\"location\\\": {\\n \\\"lat\\\": 40.7564842,\\n \\\"lng\\\": -73.9853808\\n },\\n \\\"place_id\\\": \\\"ChIJVS2qI1VYwokRFo18YsKvHYM\\\",\\n \\\"rating\\\": 4.6,\\n \\\"types\\\": [\\n \\\"establishment\\\",\\n \\\"food\\\",\\n \\\"point_of_interest\\\",\\n \\\"restaurant\\\"\\n ]\\n },\\n {\\n \\\"name\\\": \\\"La Lanterna di Vittorio\\\",\\n \\\"formatted_address\\\": \\\"129 MacDougal St, New York, NY 10012, USA\\\",\\n \\\"location\\\": {\\n \\\"lat\\\": 40.7306651,\\n \\\"lng\\\": -74.0000496\\n },\\n \\\"place_id\\\": \\\"ChIJPxCEeZFZwokRzk5UghlXZWU\\\",\\n \\\"rating\\\": 4.5,\\n \\\"types\\\": [\\n \\\"bar\\\",\\n \\\"cafe\\\",\\n \\\"establishment\\\",\\n \\\"food\\\",\\n \\\"night_club\\\",\\n \\\"point_of_interest\\\",\\n \\\"restaurant\\\",\\n \\\"store\\\"\\n ]\\n },\\n {\\n \\\"name\\\": \\\"Scarpetta\\\",\\n \\\"formatted_address\\\": \\\"88 Madison Ave, New York, NY 10016, USA\\\",\\n \\\"location\\\": {\\n \\\"lat\\\": 40.74458180000001,\\n \\\"lng\\\": -73.9856202\\n },\\n \\\"place_id\\\": \\\"ChIJt8W7n79ZwokRsyXs3QDbE3o\\\",\\n \\\"rating\\\": 4.6,\\n \\\"types\\\": [\\n \\\"bar\\\",\\n \\\"establishment\\\",\\n \\\"food\\\",\\n \\\"point_of_interest\\\",\\n \\\"restaurant\\\"\\n ]\\n },\\n {\\n \\\"name\\\": \\\"Osteria Nonnino\\\",\\n \\\"formatted_address\\\": \\\"637 Hudson St, New York, NY 10014, USA\\\",\\n \\\"location\\\": {\\n \\\"lat\\\": 40.7388972,\\n \\\"lng\\\": -74.0057401\\n },\\n \\\"place_id\\\": \\\"ChIJu3WF-z9ZwokRSIlFddmXX1k\\\",\\n \\\"rating\\\": 4.8,\\n \\\"types\\\": [\\n \\\"establishment\\\",\\n \\\"food\\\",\\n \\\"point_of_interest\\\",\\n \\\"restaurant\\\"\\n ]\\n },\\n {\\n \\\"name\\\": \\\"OLIO E PIÙ\\\",\\n \\\"formatted_address\\\": \\\"3 Greenwich Ave, New York, NY 10014, USA\\\",\\n \\\"location\\\": {\\n \\\"lat\\\": 40.7338208,\\n \\\"lng\\\": -73.99979309999999\\n },\\n \\\"place_id\\\": \\\"ChIJ_RUJvZZZwokRNUEv3K4nSik\\\",\\n \\\"rating\\\": 4.7,\\n \\\"types\\\": [\\n \\\"establishment\\\",\\n \\\"food\\\",\\n \\\"point_of_interest\\\",\\n \\\"restaurant\\\"\\n ]\\n },\\n {\\n \\\"name\\\": \\\"La Pecora Bianca SoHo\\\",\\n \\\"formatted_address\\\": \\\"54 Prince St, New York, NY 10012, USA\\\",\\n \\\"location\\\": {\\n \\\"lat\\\": 40.72346,\\n \\\"lng\\\": -73.9963118\\n },\\n \\\"place_id\\\": \\\"ChIJfe6ZYKFZwokR_mHm7GNoAxc\\\",\\n \\\"rating\\\": 4.8,\\n \\\"types\\\": [\\n \\\"establishment\\\",\\n \\\"food\\\",\\n \\\"point_of_interest\\\",\\n \\\"restaurant\\\"\\n ]\\n },\\n {\\n \\\"name\\\": \\\"Osteria Barocca\\\",\\n \\\"formatted_address\\\": \\\"133 Mulberry St, New York, NY 10013, USA\\\",\\n \\\"location\\\": {\\n \\\"lat\\\": 40.7186467,\\n \\\"lng\\\": -73.99792409999999\\n },\\n \\\"place_id\\\": \\\"ChIJK0YQvsRZwokR9r2ubFZRd7E\\\",\\n \\\"rating\\\": 4.6,\\n \\\"types\\\": [\\n \\\"establishment\\\",\\n \\\"food\\\",\\n \\\"point_of_interest\\\",\\n \\\"restaurant\\\"\\n ]\\n },\\n {\\n \\\"name\\\": \\\"Sicily Osteria\\\",\\n \\\"formatted_address\\\": \\\"328-330 W 46th St, New York, NY 10036, USA\\\",\\n \\\"location\\\": {\\n \\\"lat\\\": 40.7601601,\\n \\\"lng\\\": -73.9891863\\n },\\n \\\"place_id\\\": \\\"ChIJ03krMjJZwokRYRF15upaCHw\\\",\\n \\\"rating\\\": 4.6,\\n \\\"types\\\": [\\n \\\"establishment\\\",\\n \\\"food\\\",\\n \\\"point_of_interest\\\",\\n \\\"restaurant\\\"\\n ]\\n },\\n {\\n \\\"name\\\": \\\"La Pecora Bianca NoMad\\\",\\n \\\"formatted_address\\\": \\\"1133 Broadway, New York, NY 10010, USA\\\",\\n \\\"location\\\": {\\n \\\"lat\\\": 40.7436754,\\n \\\"lng\\\": -73.9891719\\n },\\n \\\"place_id\\\": \\\"ChIJPf0L06VZwokRD6L-7RV-7Yg\\\",\\n \\\"rating\\\": 4.7,\\n \\\"types\\\": [\\n \\\"establishment\\\",\\n \\\"food\\\",\\n \\\"point_of_interest\\\",\\n \\\"restaurant\\\"\\n ]\\n },\\n {\\n \\\"name\\\": \\\"Da Andrea Greenwich Village\\\",\\n \\\"formatted_address\\\": \\\"35 W 13th St, New York, NY 10011, USA\\\",\\n \\\"location\\\": {\\n \\\"lat\\\": 40.7361918,\\n \\\"lng\\\": -73.9958515\\n },\\n \\\"place_id\\\": \\\"ChIJS1uszZRZwokRbKPTzS6hnWM\\\",\\n \\\"rating\\\": 4.6,\\n \\\"types\\\": [\\n \\\"establishment\\\",\\n \\\"food\\\",\\n \\\"point_of_interest\\\",\\n \\\"restaurant\\\"\\n ]\\n },\\n {\\n \\\"name\\\": \\\"Via Carota\\\",\\n \\\"formatted_address\\\": \\\"51 Grove St, New York, NY 10014, USA\\\",\\n \\\"location\\\": {\\n \\\"lat\\\": 40.7331959,\\n \\\"lng\\\": -74.00364940000001\\n },\\n \\\"place_id\\\": \\\"ChIJr7FOf5NZwokRu3CDYtzVmuQ\\\",\\n \\\"rating\\\": 4.4,\\n \\\"types\\\": [\\n \\\"establishment\\\",\\n \\\"food\\\",\\n \\\"point_of_interest\\\",\\n \\\"restaurant\\\"\\n ]\\n },\\n {\\n \\\"name\\\": \\\"Masseria Dei Vini\\\",\\n \\\"formatted_address\\\": \\\"887 9th Ave, New York, NY 10019, USA\\\",\\n \\\"location\\\": {\\n \\\"lat\\\": 40.7684477,\\n \\\"lng\\\": -73.9857048\\n },\\n \\\"place_id\\\": \\\"ChIJeVfr_FhYwokRVM1IjwyvQpE\\\",\\n \\\"rating\\\": 4.6,\\n \\\"types\\\": [\\n \\\"bar\\\",\\n \\\"establishment\\\",\\n \\\"food\\\",\\n \\\"point_of_interest\\\",\\n \\\"restaurant\\\"\\n ]\\n },\\n {\\n \\\"name\\\": \\\"La Masseria NY\\\",\\n \\\"formatted_address\\\": \\\"235 W 48th St, New York, NY 10036, USA\\\",\\n \\\"location\\\": {\\n \\\"lat\\\": 40.76088499999999,\\n \\\"lng\\\": -73.9863\\n },\\n \\\"place_id\\\": \\\"ChIJjfIegVZYwokRyqdwvzBC27s\\\",\\n \\\"rating\\\": 4.5,\\n \\\"types\\\": [\\n \\\"bar\\\",\\n \\\"establishment\\\",\\n \\\"food\\\",\\n \\\"point_of_interest\\\",\\n \\\"restaurant\\\"\\n ]\\n },\\n {\\n \\\"name\\\": \\\"Zia Maria Little Italy\\\",\\n \\\"formatted_address\\\": \\\"138 Mulberry St, New York, NY 10013, USA\\\",\\n \\\"location\\\": {\\n \\\"lat\\\": 40.7187113,\\n \\\"lng\\\": -73.9975649\\n },\\n \\\"place_id\\\": \\\"ChIJm7QTo_lZwokRRnjGuJABJPQ\\\",\\n \\\"rating\\\": 4.6,\\n \\\"types\\\": [\\n \\\"bar\\\",\\n \\\"establishment\\\",\\n \\\"food\\\",\\n \\\"point_of_interest\\\",\\n \\\"restaurant\\\",\\n \\\"store\\\"\\n ]\\n },\\n {\\n \\\"name\\\": \\\"Piccola Cucina Osteria Siciliana\\\",\\n \\\"formatted_address\\\": \\\"196 Spring St, New York, NY 10012, USA\\\",\\n \\\"location\\\": {\\n \\\"lat\\\": 40.7250308,\\n \\\"lng\\\": -74.0032774\\n },\\n \\\"place_id\\\": \\\"ChIJowhlWYxZwokRgrvA8f_x8Es\\\",\\n \\\"rating\\\": 4.6,\\n \\\"types\\\": [\\n \\\"establishment\\\",\\n \\\"food\\\",\\n \\\"point_of_interest\\\",\\n \\\"restaurant\\\"\\n ]\\n },\\n {\\n \\\"name\\\": \\\"Trattoria Trecolori\\\",\\n \\\"formatted_address\\\": \\\"254 W 47th St, New York, NY 10036, USA\\\",\\n \\\"location\\\": {\\n \\\"lat\\\": 40.7599972,\\n \\\"lng\\\": -73.9867421\\n },\\n \\\"place_id\\\": \\\"ChIJYepwLVRYwokRsvXclA3XFqo\\\",\\n \\\"rating\\\": 4.5,\\n \\\"types\\\": [\\n \\\"bar\\\",\\n \\\"establishment\\\",\\n \\\"food\\\",\\n \\\"point_of_interest\\\",\\n \\\"restaurant\\\"\\n ]\\n },\\n {\\n \\\"name\\\": \\\"Don Angie\\\",\\n \\\"formatted_address\\\": \\\"103 Greenwich Ave, New York, NY 10014, USA\\\",\\n \\\"location\\\": {\\n \\\"lat\\\": 40.7378037,\\n \\\"lng\\\": -74.00209389999999\\n },\\n \\\"place_id\\\": \\\"ChIJfxC19JVZwokRX2uKgQAL17c\\\",\\n \\\"rating\\\": 4.6,\\n \\\"types\\\": [\\n \\\"establishment\\\",\\n \\\"food\\\",\\n \\\"point_of_interest\\\",\\n \\\"restaurant\\\"\\n ]\\n },\\n {\\n \\\"name\\\": \\\"Casa D'Angelo New York\\\",\\n \\\"formatted_address\\\": \\\"146 Mulberry St, New York, NY 10013, USA\\\",\\n \\\"location\\\": {\\n \\\"lat\\\": 40.7188063,\\n \\\"lng\\\": -73.9972296\\n },\\n \\\"place_id\\\": \\\"ChIJCxhnSohZwokRxmUkfrQWLfs\\\",\\n \\\"rating\\\": 4.6,\\n \\\"types\\\": [\\n \\\"bar\\\",\\n \\\"establishment\\\",\\n \\\"food\\\",\\n \\\"point_of_interest\\\",\\n \\\"restaurant\\\"\\n ]\\n }\\n ]\\n}\"]","error":null} ``` #### Example 2: Image Generation model: - gemini-3-pro-image-preview - gemini-2.5-flash-image ``` npx onekey agent gemini-nano-banana/gemini-nano-banana generate_image_gemini '{"model":"gemini-2.5-flash-image", "prompt":"Generate a minecraft scene of steve fighting zombies in purple crystal fields."}' ``` ``` {"success":true,"data":"[\"{\\n \\\"image_path\\\": \\\"/data/python/mcp_tool_use/files-wd/user_2177/7629e198-b10c-4a8b-8f61-5c0a6744cab0/gemini_output_images.png\\\",\\n \\\"image_url\\\": \\\"https://us-static.aiagenta2z.com/local/files-wd/user_2177/7629e198-b10c-4a8b-8f61-5c0a6744cab0/gemini_output_images.png\\\",\\n \\\"message\\\": \\\"\\\",\\n \\\"success\\\": true\\n}\"]","error":null} ## url of image generated https://us-static.aiagenta2z.com/local/files-wd/user_2177/7629e198-b10c-4a8b-8f61-5c0a6744cab0/gemini_output_images.png ``` **Return Image URL Result** Image generation #### Example 3: 3D Generation Step By Step Assembly Charts Build Plans prompt: ```bash export DEEPNLP_ONEKEY_ROUTER_ACCESS=YOUR_ACCESS_KEY npx onekey agent craftsman-agent/craftsman-agent generate_lego_build_plan '{"prompt":"Build Lego yacht with 5 decks using blue and white bricks","images":[],"mode":"basic"}' --timeout 60000 ``` 3D Step by Step Assembly Charts Image generation ### Example 4. Generate Tesla Car Wrap prompt: model YL F1 race car blue and purple with stars model: tesla_model_yl ```bash export DEEPNLP_ONEKEY_ROUTER_ACCESS=YOUR_ACCESS_KEY npx onekey agent craftsman-agent/craftsman-agent generate_tesla_wraps '{"prompt":"I would like to paint my tesla model YL similar to F1 race car, color of a blue and purple with stars","images":[],"mode":"basic","car_model":"tesla_model_yl","output_number":1}' --timeout 60000 ``` Tesla Model YL ``` {"success":true,"text":"","images":[{"url":"https://us-static.aiagenta2z.com/local/files-wd/onekey_llm_router/722aa763-ec48-48b5-acfd-c569f89ffc50.png"}]} ``` ## Tutorial ## 1. OneKey Agent Router Usage ### **CLI** ```bash npx onekey agent [--timeout ] npx onekey mcp [--name config_name] [--timeout ] npx onekey llm --provider --model --messages [--temperature ] [--response-format ] [--options ] [--timeout ] npx onekey llm --payload [--timeout ] ``` Supported ids can be found in the table below. ### **Skills** You can download skills (wrapper of OneKey Gateway in CLI with SKILL.md) to your local agents, such as claude code, codex, gemini, openclaw. `agtm` package to download skills ```bash npx agtm skills add https://github.com/aiagenta2z/onekey-gateway ## add all onekey router skills npx agtm skills add aiagenta2z/onekey-gateway --skill google-maps -g ## install to global path npx agtm skills add aiagenta2z/onekey-gateway --skill gemini-nano-banana ``` `skills` package ``` npx skills add https://github.com/aiagenta2z/onekey-gateway ## single skill npx skills add https://github.com/aiagenta2z/onekey-gateway --skill amap-maps-streamableHTTP ``` ### **API** Set env variable and use curl to get Agent APIs of Google maps ``` export DEEPNLP_ONEKEY_ROUTER_ACCESS=your_access_key curl -v -X POST "https://agent.deepnlp.org/agent_router" \ -H "Content-Type: application/json" \ -H "X-OneKey: $DEEPNLP_ONEKEY_ROUTER_ACCESS" \ -d '{"unique_id":"google-maps/google-maps","api_id":"maps_geocode","data":{"address": "Times Square, New York"}}' ``` ## 2. OneKey MCP Router ### **CLI** ```bash npx onekey mcp [--name config_name] [--timeout ] ``` Use `npx onekey mcp` cli to generate the mcp config and copy paste to your clients. ```bash npx onekey mcp google-maps ``` Put below config into your clients ```shell { "mcpServers": { "deepnlp-onekey-google-maps": { "url": "https://agent.deepnlp.org/mcp?server_name=google-maps&onekey=your_access_key" } } } ``` ## 3. OneKey LLM Router Call LLM API, such as Gemini 3 Flash Models, Image Generation Model Nano Banana ### **CLI** ```bash npx onekey llm --provider --model --messages [--temperature ] [--response-format ] [--options ] npx onekey llm --payload ``` ```bash npx onekey llm --provider gemini --model gemini-3-flash-preview --messages @messages.json --temperature 0.7 --response-format json --timeout 60000 ``` ## Gateway Support Summary | Domain | Unique Id | Skill ID | API ID | |----------------------|-----------------------------------------------------------------------------------------------------------------|--------------------------|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------| | Map (Google) | google-maps/google-maps [Doc](docs/onekey-gateway/google-maps/README.md) | google-maps | maps_directions
maps_distance_matrix
maps_elevation
maps_geocode
maps_place_details
maps_reverse_geocode
maps_search_places | | Map (Amap/Gaode Map) | amap-maps-streamableHTTP/amap-maps-streamableHTTP [Doc](docs/onekey-gateway/amap-maps-streamableHTTP/README.md) | amap-maps-streamableHTTP | maps_around_search
maps_direction_bicycling
maps_direction_driving
maps_direction_transit_integrated
maps_direction_walking
maps_distance
maps_geo
maps_ip_location
maps_regeocode
maps_schema_navi
maps_schema_personal_map
maps_schema_take_taxi
maps_search_detail
maps_text_search
maps_weather | | Map (Baidu) | baidu-maps-sse/baidu-maps-sse [Doc](docs/onekey-gateway/baidu-maps-sse/README.md) | baidu-maps-sse | maps_directions
maps_distance_matrix
maps_elevation
maps_geocode
maps_place_details
maps_reverse_geocode
maps_search_places | | Image Search | bing-image-search-mcp/bing-image-search-mcp [Doc](docs/onekey-gateway/bing-image-search-mcp/README.md) | bing-image-search-mcp | search_images
search_images_batch | | Web Search | brave-search/brave-search [Doc](docs/onekey-gateway/brave-search/README.md) | brave-search | brave_local_search
brave_web_search | | Web Search | google-search/google-search [Doc](docs/onekey-gateway/google-search/README.md) | google-search | google_search | | Web Search | tavily-remote-mcp/tavily-remote-mcp [Doc](docs/onekey-gateway/tavily-remote-mcp/README.md) | tavily-remote-mcp | tavily_crawl
tavily_extract
tavily_map
tavily_research
tavily_search | | Web Crawl / Browser | firecrawl-mcp/firecrawl-mcp [Doc](docs/onekey-gateway/firecrawl-mcp/README.md) | firecrawl-mcp | firecrawl_agent
firecrawl_agent_status
firecrawl_browser_create
firecrawl_browser_delete
firecrawl_browser_execute
firecrawl_browser_list
firecrawl_check_crawl_status
firecrawl_crawl
firecrawl_extract
firecrawl_map
firecrawl_scrape
firecrawl_search | | Image Generation | gemini/gemini [Doc](docs/onekey-gateway/gemini/README.md) | gemini (Nano Banana) | generate_image_gemini
generate_image_nano_banana
generate_image_nano_banana_with_reference
ocr_extract_text_from_image
list_items_from_image | | Image Generation | gemini-nano-banana/gemini-nano-banana [Doc](docs/onekey-gateway/gemini-nano-banana/README.md) | gemini-nano-banana | generate_image_gemini
generate_image_nano_banana
generate_image_nano_banana_with_reference
ocr_extract_text_from_image
list_items_from_image | | Charts | mcp-server-chart/mcp-server-chart [Doc](docs/onekey-gateway/mcp-server-chart/README.md) | mcp-server-chart | generate_area_chart
generate_bar_chart
generate_boxplot_chart
generate_column_chart
generate_district_map
generate_dual_axes_chart
generate_fishbone_diagram
generate_flow_diagram
generate_funnel_chart
generate_histogram_chart
generate_line_chart
generate_liquid_chart
generate_mind_map
generate_network_graph
generate_organization_chart
generate_path_map
generate_pie_chart
generate_pin_map
generate_radar_chart
generate_sankey_chart
generate_scatter_chart
generate_spreadsheet
generate_treemap_chart
generate_venn_chart
generate_violin_chart
generate_waterfall_chart
generate_word_cloud_chart | | Research QA | perplexity/perplexity [Doc](docs/onekey-gateway/perplexity/README.md) | perplexity | perplexity_ask
perplexity_reason
perplexity_research
perplexity_search | | 3D Generation | craftsman-agent/craftsman-agent [Doc](docs/onekey-gateway/craftsman-agent/README.md) | craftsman-agent | generate_lego_build_plan
generate_minecraft_build_plan
generate_tesla_wraps | | Finance | aiagenta2z/financeagent [Doc](docs/onekey-gateway/financeagent/README.md) | financeagent | get_hk_stock_market_hkex
get_cn_stock_market_shanghai_shenzhen
get_us_stock_market_nyse_nasdaq_dow
get_uk_stock_market_lse
get_india_stock_market_nse_india | | Healthcare | fdcnal/usda-fooddata-central-agent [Doc](docs/onekey-gateway/healthcareagent/README.md) | usda-fooddata-central-agent | get_foods_multiple
get_food_by_id
list_foods
search_foods | | Healthcare | aiagenta2z/healthcaregent [Doc](docs/onekey-gateway/usda-fooddata-central-agent/README.md) | healthcareagent | | #### Example 1: Google Maps Search ```bash npx onekey agent google-maps/google-maps maps_search_places '{"query":"New York City Italian Restaurants"}' ``` ```bash curl -X POST "https://agent.deepnlp.org/agent_router" \ -H "Content-Type: application/json" \ -H "X-OneKey: $DEEPNLP_ONEKEY_ROUTER_ACCESS" \ -d '{ "unique_id": "google-maps/google-maps", "api_id": "maps_search_places", "data": { "query": "New York City Italian Restaurants" } }' ``` #### Example 2: Image Generation model: - gemini-3-pro-image-preview - gemini-2.5-flash-image ``` npx onekey agent gemini-nano-banana/gemini-nano-banana generate_image_gemini '{"model":"gemini-2.5-flash-image", "prompt":"Generate a minecraft scene of steve fighting zombies in purple crystal fields."}' ``` ``` {"success":true,"data":"[\"{\\n \\\"image_path\\\": \\\"/data/python/mcp_tool_use/files-wd/user_2177/7629e198-b10c-4a8b-8f61-5c0a6744cab0/gemini_output_images.png\\\",\\n \\\"image_url\\\": \\\"https://us-static.aiagenta2z.com/local/files-wd/user_2177/7629e198-b10c-4a8b-8f61-5c0a6744cab0/gemini_output_images.png\\\",\\n \\\"message\\\": \\\"\\\",\\n \\\"success\\\": true\\n}\"]","error":null} ## Return Image url: https://us-static.aiagenta2z.com/local/files-wd/user_2177/7629e198-b10c-4a8b-8f61-5c0a6744cab0/gemini_output_images.png ``` ## Supported Agents ### Codex
Onekey Gateway Usage #### Install ``` npm -g install @aiagenta2z/onekey-gateway npx agtm skills add https://github.com/aiagenta2z/onekey-gateway -a codex npx skills add https://github.com/aiagenta2z/onekey-gateway ``` #### Setup Env File ``` vim .env DEEPNLP_ONEKEY_ROUTER_ACCESS=${your_accesskey} ``` Export Env ``` export DEEPNLP_ONEKEY_ROUTER_ACCESS=your_accesskey ``` #### Run Codex ``` codex ## Show Skills Loaded /skills ## Example 1. Generate a image poster academic/business/professional styles /skills Read the README file and relevant project files to understand the project. Then use Gemini Nano Banana to generate a high-quality image (poster) that visually represents the project in a professional, academic, or business style and output to local folder /docs/*.png ### Example 2. Call Google Maps to Search places /skills Use Google Maps Search Best Italian Restaurants in New York ``` #### Results ``` • Updated Plan └ ✔ Scan README and project structure ✔ Summarize project goals and features ✔ Draft poster concept and prompt □ Generate poster image via Gemini □ Save PNG to docs folder • Ran npx --yes onekey agent gemini-nano-banana/gemini-nano-banana generate_image_gemini @docs/poster_request_gemini.json └ {"success":true,"data":[{"image_path":"/data/python/mcp_tool_use/files-wd/user_1808/9f03533e-af88-4959-b6b4-0131230728b9/Users/rockingdingo/Desktop/workspace/ github/aiagenta2z/onekey-gateway-test/docs/onekey_gateway_poster_gemini.png","image_url":"https://us-static.aiagenta2z.com/local/files-wd/ user_1808/9f03533e-af88-4959-b6b4-0131230728b9/Users/rockingdingo/Desktop/workspace/github/aiagenta2z/onekey-gateway-test/docs/ onekey_gateway_poster_gemini.png","message":"Here is the OneKey Gateway poster you requested.","success":true}],"error":null} ```
### Gemini-CLI
Onekey Gateway Usage #### Install ``` npm -g install @aiagenta2z/onekey-gateway npx agtm skills add https://github.com/aiagenta2z/onekey-gateway -a gemini-cli npx skills add https://github.com/aiagenta2z/onekey-gateway ``` ### Setup Env File ``` vim .env DEEPNLP_ONEKEY_ROUTER_ACCESS=${your_accesskey} ``` Export Env ``` export DEEPNLP_ONEKEY_ROUTER_ACCESS=your_accesskey ``` #### Run Gemini-CLI ``` gemini ## List if the onekey skills are loaded /skills list ## Generate a business/academic/professional poster of the architecture of the project Read the README file and relevant project files to understand the project. Then use Onekey Gateway skills of Gemini Nano Banana to generate a high-quality image (poster) that visually represents the project in a professional, academic, or business style and output to local folder /docs/*.png ### Call Google Maps to Search places Use onekey gateway skills Google Maps Search Best Italian Restaurants in New York ```
### Claude Code
Onekey Gateway Usage ```shell npx agtm skills add https://github.com/aiagenta2z/onekey-gateway -a claude-code claude > Use Nano Banana to Generate an Image of academic poster of the project > Search top italian restaurants and add to calendars. ```
### OpenClaw
Onekey Gateway Usage ```shell npx agtm skills add https://github.com/aiagenta2z/onekey-gateway -a openclaw > Use Nano Banana to Generate an Image of academic poster of the project > Search top italian restaurants and add to calendars. ```
### Related [OneKey Gateway Documentation](https://www.deepnlp.org/doc/onekey_gateway) [AI Agent Marketplace](https://www.deepnlp.org/store/ai-agent) [GitHub Agtm](https://github.com/aiagenta2z/agtm) [Submit AI Agent](https://www.deepnlp.org/workspace/my_ai_services) [Agent Infra Deployment](https://www.deepnlp.org/workspace/deploy) [A2Z Payment SDK](https://www.deepnlp.org/agent/agent-a2z-payment)