Obscurify AI
OfficialSummary: This MCP server gives an AI agent private, pay-per-use access to 200+ chat models (Claude, GPT, Gemini, Llama, DeepSeek, Qwen) plus Stable Diffusion image generation through Obscurify.ai.
ask— send one prompt to any model (with optional system prompt,max_tokens,temperature,reasoning_effort,show_reasoning); returns the reply plus model, tokens and cost.chat— multi-turn conversation by passing the wholemessageshistory (the server holds no state); returns the next assistant reply.second_opinion— the same prompt sent to 2–5 models in parallel, with answers side by side for comparison, fact-checking or tie-breaking.list_models— browse model IDs with context window, input/output price, tool support and required plan; filter by name, plan, price, context, tools or output price. Works without an API key.generate_image— text-to-image with Stable Diffusion models (SDXL, SD 1.5, LCM), with size, seed, negative prompt;list: trueshows image models and prices.balance— check balance, plan, usage this billing period, and whether frontier and uncensored models are available.Privacy — text requests use zero-data-retention routes; the server is a thin client with no telemetry, logging or file access.
Setup — runs via
npx -y @obscurify/mcp(or Docker) over stdio, or self-hosted stateless Streamable HTTP where each request can carry its own bearer key.
Routes prompts to Meta's Llama models through Obscurify.ai — e.g. meta-llama/llama-3.3-70b-instruct, which is also part of the default second_opinion panel for comparing answers across models.
Routes prompts to OpenAI's GPT models through Obscurify.ai — call ask, chat or second_opinion with a model such as openai/gpt-5.5 and get the reply plus model, token usage and cost.
Obscurify.ai MCP server
Give your AI agent private access to 200+ other models. This Model Context Protocol server lets Claude Desktop, Claude Code, Cursor, VS Code, Codex, opencode and any other MCP client ask Claude, GPT, Gemini, Llama, DeepSeek, Qwen and more through Obscurify.ai.
Obscurify.ai is a privacy-focused, OpenAI-compatible AI platform. Text requests only use zero-data-retention routes, prompts and responses aren't stored on its servers, and accounts are a random 16-digit account number with no email, phone or password.
Tools
Tool | What it does |
| One prompt to any model, with an optional system prompt, |
| Multi-turn: pass the whole |
| The same prompt to 2-5 models in parallel, every answer side by side. Good for checking a claim, reviewing code or a plan, or breaking a tie. |
| Model IDs with context window, price per million input/output tokens, tool support and the plan each needs. Filter by name, plan, price or tool support. Works without a key. |
| Balance, plan, usage this billing period, and whether the account can use frontier and uncensored models. |
Defaults are cheap and fast: ask and chat use deepseek/deepseek-v4-flash, and second_opinion asks deepseek/deepseek-v4-flash, meta-llama/llama-3.3-70b-instruct and qwen/qwen3.8-27b. Pass model / models to use anything else, e.g. anthropic/claude-sonnet-5.5 or openai/gpt-5.5.
Related MCP server: Gemini CLI MCP Server
Get an API key
Create an account at obscurify.ai. You get a random account number: no email, no password.
Add a balance (card, Monero, Bitcoin or Lightning). The API works on every plan with a positive balance. Frontier models need the Casual plan or higher, and uncensored models need Pro (pricing).
On obscurify.ai/account, click Show API Key and copy the
sk-obs1_...key.
Calls cost money from your balance at each model's listed price (model list). Every reply shows what it cost.
Setup
The server runs with npx -y @obscurify/mcp (Node.js 20 or newer) and reads the key from OBSCURIFY_API_KEY.
Claude Code
claude mcp add obscurify -e OBSCURIFY_API_KEY=sk-obs1_... -- npx -y @obscurify/mcpAdd -s user to make it available in every project.
Claude Desktop
Settings → Developer → Edit Config, then add to claude_desktop_config.json:
{
"mcpServers": {
"obscurify": {
"command": "npx",
"args": ["-y", "@obscurify/mcp"],
"env": { "OBSCURIFY_API_KEY": "sk-obs1_..." }
}
}
}Restart Claude Desktop.
Cursor
Add the same mcpServers block to ~/.cursor/mcp.json (all projects) or .cursor/mcp.json (one project):
{
"mcpServers": {
"obscurify": {
"command": "npx",
"args": ["-y", "@obscurify/mcp"],
"env": { "OBSCURIFY_API_KEY": "sk-obs1_..." }
}
}
}VS Code (GitHub Copilot agent mode)
Create .vscode/mcp.json. VS Code asks for the key once and stores it securely:
{
"inputs": [
{ "type": "promptString", "id": "obscurify-key", "description": "Obscurify.ai API key", "password": true }
],
"servers": {
"obscurify": {
"type": "stdio",
"command": "npx",
"args": ["-y", "@obscurify/mcp"],
"env": { "OBSCURIFY_API_KEY": "${input:obscurify-key}" }
}
}
}Codex CLI
codex mcp add obscurify --env OBSCURIFY_API_KEY=sk-obs1_... -- npx -y @obscurify/mcpor in ~/.codex/config.toml:
[mcp_servers.obscurify]
command = "npx"
args = ["-y", "@obscurify/mcp"]
env = { OBSCURIFY_API_KEY = "sk-obs1_..." }
# Optional: don't ask before each paid call (ask, chat, second_opinion)
default_tools_approval_mode = "approve"The tools that spend from your balance are marked as not read-only, so Codex asks before running them unless you set default_tools_approval_mode.
opencode
opencode mcp add obscurify --env OBSCURIFY_API_KEY=sk-obs1_... -- npx -y @obscurify/mcpThat writes this to opencode.json (add --global for your global config):
{
"mcp": {
"servers": {
"obscurify": {
"type": "local",
"command": ["npx", "-y", "@obscurify/mcp"],
"environment": { "OBSCURIFY_API_KEY": "sk-obs1_..." }
}
}
}
}Windsurf, Cline, Zed and other clients
Use the command npx -y @obscurify/mcp with the environment variable OBSCURIFY_API_KEY in the client's MCP settings.
Docker
docker build -t obscurify-mcp https://github.com/Obscurify-ai/obscurify-mcp.git
docker run -i --rm -e OBSCURIFY_API_KEY=sk-obs1_... obscurify-mcpStreamable HTTP
For clients that connect over HTTP, run the server yourself:
OBSCURIFY_API_KEY=sk-obs1_... npx -y @obscurify/mcp --http --port 3000
# MCP endpoint: http://127.0.0.1:3000/mcp health: /healthIt binds to 127.0.0.1 unless you pass --host. Each request may carry its own key as Authorization: Bearer sk-obs1_..., which takes precedence over OBSCURIFY_API_KEY, so one server can serve several accounts. The HTTP mode is stateless.
Configuration
Variable | Default | |
| (none) | Your |
|
| Model for |
|
| Comma-separated models for |
|
| API base URL. |
|
| Per-request timeout. |
| off |
|
|
| Where the private-session proxy listens. |
|
| The proxy binary, if it isn't on |
Private sessions
Set OBSCURIFY_PRIVATE=1 and ask, chat and second_opinion go through private sessions (unlinkable inference): each conversation is paid from a wallet of blind-signed vouchers on your machine, through a session with its own random secret, so Obscurify.ai can't tell which account a question came from or which questions belong together. Your API key only tops the wallet up; it never goes with a question.
It needs the open-source proxy: cargo install unlinkable-inference-cli. The MCP server starts it when it isn't running (and leaves it running, so sessions close on their own when idle). list_models and balance stay as they are: the model list is public and the balance belongs to the account.
"env": { "OBSCURIFY_API_KEY": "sk-obs1_...", "OBSCURIFY_PRIVATE": "1" }Guide for every client (Claude Code, Codex, pi, SDKs): obscurify.ai/tutorials/private_sessions.
Example prompts
"Ask Gemini and GPT what they think of this migration plan, then summarise where they disagree." (
second_opinion)"Find the cheapest model with tool calling and a 1M context." (
list_models)"Have Llama 3.3 70B rewrite this paragraph for a general audience." (
ask)"How much have I spent on Obscurify this month?" (
balance)
Privacy
The server is a thin client: it sends your prompts to https://obscurify.ai with your key and nothing else. No telemetry, no logging to disk, no file access. What Obscurify.ai keeps is described in its privacy policy: prompts and responses are not stored, and identities are scrambled before requests reach model providers. Providers still see the prompt text, so don't send secrets you wouldn't send to the model itself.
Tested
On 2026-10-28 with a Pay As You Go key, every tool was called from MCP Inspector 2.12 (CLI), Claude Code 2.1, Codex CLI 0.165 (both from the local build and from npx -y @obscurify/mcp), and the SDK client in scripts/e2e.mjs over stdio and Streamable HTTP. opencode 2.1 connects and lists the tools.
On 2026-11-15, 0.3.0's private mode passed scripts/e2e.mjs over stdio against obscurify.ai (every tool, questions through the proxy's sessions) and against a local server that had to start the proxy itself.
Development
npm install
npm run build
OBSCURIFY_API_KEY=sk-obs1_... node scripts/e2e.mjs stdio # live end-to-end test, costs well under a cent
OBSCURIFY_API_KEY=sk-obs1_... node scripts/e2e.mjs http
npx @modelcontextprotocol/inspector node dist/index.jsReleases: bump the version in package.json and server.json, then push a vX.Y.Z tag. The workflow publishes to npm with provenance and to the MCP Registry.
Links
Website: obscurify.ai
API docs: obscurify.ai/api-docs
Contact: contact@obscurify.ai
License
MIT
Available Tools
5 toolsaskAsk a modelA
Send one prompt to any Obscurify.ai model and get its reply. Useful for delegating a subtask to a cheaper or specialised model, getting an outside view, or using a model you don't otherwise have. Prompts go through zero-data-retention routes and are not stored by Obscurify.ai. Default model: deepseek/deepseek-v4-flash (fast and cheap). Use list_models to find other model IDs. For follow-ups that need earlier turns use chat; to compare several models use second_opinion.
Inputs: prompt (required), model, system, max_tokens, +3 more.
balanceAccount balance and planARead-only
Show the Obscurify.ai account's balance (USD), plan, included credit used this billing period, token usage, and whether it can use frontier and uncensored models. Use it before a large batch of calls, or when a call fails with a balance (402) or plan (403) error.
Inputs: no named parameters.
chatMulti-turn chatA
Continue a multi-turn conversation with an Obscurify.ai model. Pass the whole history in messages (the server keeps no state) and get the next assistant reply. For a single question, ask is simpler. Default model: deepseek/deepseek-v4-flash.
Inputs: messages (required), model, max_tokens, temperature, +2 more.
list_modelsList modelsARead-only
List Obscurify.ai chat models with their IDs, context window, price per million input/output tokens (USD), tool-calling support and the plan they need (any = every account with a balance; casual = Casual or Pro; pro = Pro only). Filter by name, plan, price or tool support. Use it to find a model ID for ask, chat or second_opinion, or the cheapest model with what you need. Doesn't need an API key.
Inputs: plan, sort, limit, search, +2 more.
second_opinionSecond opinion from several modelsA
Send the same prompt to 2-5 different models at once and return every answer side by side, so you can compare them, check a claim, review code or a plan, or break a tie. Default models: deepseek/deepseek-v4-flash, meta-llama/llama-3.3-70b-instruct, qwen/qwen3.8-27b.
Inputs: prompt (required), models, system, max_tokens, +1 more.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 5 tools
Each tool targets a distinct mode of interaction: single prompt (ask), multi-turn (chat), multi-model comparison (second_opinion), model discovery (list_models), and account state (balance). The descriptions explicitly route agents between the overlapping prompt-sending tools ("For follow-ups ... use chat; to compare ... use second_opinion"), eliminating misselection risk.
All names are lowercase snake_case and short, which is consistent and readable. The pattern is slightly mixed in form (bare verbs like ask/chat, a noun like balance, a verb_noun like list_models, and a compound noun second_opinion), but each name still signals its action clearly.
Five tools is well-scoped for a model-gateway server: query, converse, compare, discover, and check credit. Every tool maps to a real workflow and none appear redundant or filler.
The core lifecycle (discover models -> send prompts -> converse -> compare -> verify balance/plan) is fully covered with no dead ends. Minor gaps exist: no per-model detail lookup beyond list_models, no usage history or call-log inspection, and no way to surface tool-calling behavior despite models advertising it.
Maintenance
Related MCP Connectors
Chat, image, video, and audio models from many vendors behind one remote MCP server and one key.
Talk to your public-facing AI from any MCP client — Claude, ChatGPT, Cursor, Cline, Windsurf.
Connect MCP clients to 2,000+ AI models without managing provider API keys.
Generate images, video, audio and short films with 140+ AI models from any MCP client.
Related MCP Servers
- AlicenseCqualityFmaintenanceSend requests to OpenAI, MistralAI, Anthropic, xAI, or Google AI using MCP protocol via tool or predefined prompts. Vendor API key required140MIT
- FlicenseNot gradedqualityDmaintenanceBridges Google's Gemini CLI with MCP-compatible clients and integrates OpenRouter API for access to 400+ AI models. Provides 33 specialized tools enabling multi-AI workflows, collaborations, and debates between Claude, Gemini, and other LLMs.142-
- AlicenseNot gradedqualityCmaintenanceAccess 400+ generative AI models directly from your AI assistant — generate images (FLUX, Midjourney, GPT-4o), create videos (Veo3, Kling), make music (Suno), and enhance photos, all via a single MCP server.14MIT
- AlicenseAqualityCmaintenanceHosted multi-model AI media + chat MCP server. Generates images, video, audio, face-swaps and talking-avatars, and chats across 300+ models (Claude, GPT, Gemini, DeepSeek…) - all from one balance and one API key.16MIT