SeeMe MaaS Documentation

SeeMe MaaS is a unified AI model gateway that exposes 200+ models from OpenAI, Anthropic, Google, DeepSeek and more through a single OpenAI-compatible endpoint. This guide gets you integrated in minutes.

OpenAI CompatiblePay-per-Token200+ ModelsSmart Routing

Quick Start

Register an account

Go to the login page and create an account. New accounts receive starter credits to begin calling the API immediately.

Create an API key

Once logged in, head to API Keys and click “New Key” to generate a token. Use separate keys per project so usage and limits stay clean.

Configure the base URL

Use the following URL as the Base URL in any OpenAI-compatible SDK or client:

http://115.190.128.17:8080/v1Copy

Make your first call

Swap the placeholder key for the one you just generated and run any of the snippets in the API Examples section below.

API Examples

SeeMe MaaS is fully OpenAI-compatible. Swap the base URL and API key, and your existing OpenAI client just works.

1curl http://115.190.128.17:8080/v1/chat/completions \
2  -H "Content-Type: application/json" \
3  -H "Authorization: Bearer sk-myrouter-xxxxxxxxxxxxxxxx" \
4  -d '{
5    "model": "anthropic/claude-sonnet-4-5",
6    "messages": [
7      {"role": "user", "content": "Hello, SeeMe MaaS!"}
8    ]
9  }'

Multimodal & Video

Beyond chat, SeeMe MaaS exposes OpenAI-compatible endpoints for vision understanding, image generation, image editing, and text-to-video. All endpoints share the same base URL and API key.

Image Input (Gemini / GPT-4o / Claude)Vision

In chat/completions, upgrade content to an array mixing text and image_url parts. url accepts either a public URL or an inline data:image/...;base64 string.

1curl http://115.190.128.17:8080/v1/chat/completions \
2  -H "Content-Type: application/json" \
3  -H "Authorization: Bearer sk-myrouter-xxxxxxxxxxxxxxxx" \
4  -d '{
5    "model": "google/gemini-2.5-pro",
6    "messages": [
7      {
8        "role": "user",
9        "content": [
10          {"type": "text", "text": "What is in this image?"},
11          {
12            "type": "image_url",
13            "image_url": {"url": "https://example.com/photo.jpg"}
14          }
15        ]
16      }
17    ]
18  }'

Recommended models: google/gemini-2.5-pro · openai/gpt-4o · anthropic/claude-sonnet-4-5

Image Generation (/v1/images/generations)Text-to-Image

Generate images from a text prompt. Responses contain data[].url (or b64_json for some models).

1curl http://115.190.128.17:8080/v1/images/generations \
2  -H "Content-Type: application/json" \
3  -H "Authorization: Bearer sk-myrouter-xxxxxxxxxxxxxxxx" \
4  -d '{
5    "model": "google/imagen-3",
6    "prompt": "Cyberpunk Kyoto at night, neon rain",
7    "size": "1024x1024",
8    "n": 1
9  }'

Recommended models: openai/dall-e-3 · google/imagen-3 · black-forest-labs/flux-pro

Image Edit (/v1/images/edits)Image Edit

Upload the source image (optional mask) as multipart/form-data, with a prompt describing the edit.

1curl http://115.190.128.17:8080/v1/images/edits \
2  -H "Authorization: Bearer sk-myrouter-xxxxxxxxxxxxxxxx" \
3  -F image=@source.png \
4  -F mask=@mask.png \
5  -F model=openai/gpt-image-1 \
6  -F prompt="Replace the background with a starry sky and aurora" \
7  -F size=1024x1024

Note: the request must be multipart/form-data. Do not set Content-Type manually — let the SDK or curl generate the boundary.

Text-to-Video (Submit + Poll)Text-to-Video

Video generation is asynchronous: POST /v1/videos/generations to receive a task id, then poll GET /v1/videos/generations/{id} until the task completes. Typical latency is 30 s – a few minutes.

Step 1: Submit the task

1curl http://115.190.128.17:8080/v1/videos/generations \
2  -H "Content-Type: application/json" \
3  -H "Authorization: Bearer sk-myrouter-xxxxxxxxxxxxxxxx" \
4  -d '{
5    "model": "google/veo-2",
6    "prompt": "Cinematic slow-motion waves crashing on rocks",
7    "duration": 5,
8    "aspect_ratio": "16:9"
9  }'
10# => {"id":"task_abc123","status":"pending"}

Step 2: Poll for completion

1# Poll every few seconds until status is done or failed
2curl http://115.190.128.17:8080/v1/videos/generations/task_abc123 \
3  -H "Authorization: Bearer sk-myrouter-xxxxxxxxxxxxxxxx"
4
5# {
6#   "id": "task_abc123",
7#   "status": "done",
8#   "video_url": "https://cdn.example.com/out.mp4",
9#   "thumbnail_url": "https://cdn.example.com/out.jpg"
10# }

Recommended models: google/veo-2 · kling/kling-v1 · minimax/video-01. Supported parameters (duration / aspect_ratio / fps) vary per provider — check the Model Library for details.

Supported Tools

These mainstream AI clients all work with OpenAI- or Claude-compatible APIs. Replace sk-myrouter-xxxxxxxxxxxxxxxx in the snippets with the key you generated in API Keys.

Claude CodeCLI

Anthropic’s official AI coding assistant. A single env file routes all requests through SeeMe MaaS.

Step 1: Edit `~/.claude/settings.json`

On Windows the path is %USERPROFILE%\.claude\settings.json. Paste:

1{
2  "env": {
3    "ANTHROPIC_BASE_URL": "http://115.190.128.17:8080",
4    "ANTHROPIC_AUTH_TOKEN": "sk-myrouter-xxxxxxxxxxxxxxxx",
5    "API_TIMEOUT_MS": "600000",
6    "CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC": "1",
7    "ANTHROPIC_MODEL": "anthropic/claude-sonnet-4-5",
8    "ANTHROPIC_SMALL_FAST_MODEL": "anthropic/claude-haiku-4-5",
9    "ANTHROPIC_DEFAULT_SONNET_MODEL": "anthropic/claude-sonnet-4-5",
10    "ANTHROPIC_DEFAULT_OPUS_MODEL": "anthropic/claude-sonnet-4-5",
11    "ANTHROPIC_DEFAULT_HAIKU_MODEL": "anthropic/claude-haiku-4-5"
12  }
13}

Step 2: Skip the onboarding flow

Edit ~/.claude.json (Windows: %USERPROFILE%\.claude.json) and add:

1{
2  "hasCompletedOnboarding": true
3}

Now run claude and Claude Code will route all calls through SeeMe MaaS.

Codex (OpenAI)CLI

OpenAI’s official coding agent. Point OPENAI_BASE_URL at SeeMe MaaS and you’re done.

1# Add to ~/.bashrc or ~/.zshrc
2export OPENAI_BASE_URL=http://115.190.128.17:8080/v1
3export OPENAI_API_KEY=sk-myrouter-xxxxxxxxxxxxxxxx
4
5# Launch Codex
6codex

CursorIDE

AI-first code editor with built-in chat and completions.

Open model settings

Cursor → Settings → Models

Fill in your API key

Paste your token into the OpenAI API Key field:

sk-myrouter-xxxxxxxxxxxxxxxxCopy

Override Base URL

Enable “Override OpenAI Base URL” and use:

http://115.190.128.17:8080/v1Copy

Add and enable models

Add models like openai/gpt-4o or anthropic/claude-sonnet-4-5 under Model Names, click Verify, and use them in Chat or Composer.

ContinueVS Code Plugin

Open-source AI coding assistant for VS Code and JetBrains.

Install the extension

Search “Continue” in the VS Code marketplace and install.

Edit the config file

Open ~/.continue/config.json and update the models array:

1{
2  "models": [
3    {
4      "title": "GPT-4o (SeeMe MaaS)",
5      "provider": "openai",
6      "model": "openai/gpt-4o",
7      "apiBase": "http://115.190.128.17:8080/v1",
8      "apiKey": "sk-myrouter-xxxxxxxxxxxxxxxx"
9    },
10    {
11      "title": "Claude Sonnet 4.5 (SeeMe MaaS)",
12      "provider": "openai",
13      "model": "anthropic/claude-sonnet-4-5",
14      "apiBase": "http://115.190.128.17:8080/v1",
15      "apiKey": "sk-myrouter-xxxxxxxxxxxxxxxx"
16    }
17  ]
18}

Start chatting

Press Cmd/Ctrl + L to open the chat panel and pick a model.

ClineVS Code Plugin

Autonomous coding agent that reads, writes, and runs commands for you.

Install Cline

Search “Cline” in the VS Code marketplace and install.

Configure the API

Open the Cline sidebar → settings icon → fill in:

API Provider：OpenAI Compatible

Base URL：http://115.190.128.17:8080/v1

API Key：sk-myrouter-xxxxxxxxxxxxxxxx

Model ID：anthropic/claude-sonnet-4-5

Give it a task

Describe your task in the Cline panel and let the agent take it from there.

General Desktop ClientsDesktop

All of the following speak the OpenAI-compatible protocol — configuration is largely identical.

1# Settings → Model Services → Add Provider
2Type:        OpenAI Compatible
3Name:        SeeMe MaaS
4API URL:     http://115.190.128.17:8080/v1
5API Key:     sk-myrouter-xxxxxxxxxxxxxxxx
6# After saving, click “Fetch Models” to auto-load the list.

Model List

SeeMe MaaS aggregates 200+ models from 20+ providers including OpenAI, Anthropic, Google, DeepSeek, Meta, Alibaba, Mistral and more. Browse the catalog and live pricing in the Model Library.

Browse Model Library

FAQ

Still stuck? Reach us via the channels in the footer.