LIVE · api.prxy.monster · BYOK prxy.monster

Your agent stops forgetting.
Your bill stops climbing.

prxy.monster sits in front of Claude, OpenAI, or Bedrock and runs every request through twelve modules that fix the things that keep breaking. Bring your own provider key. One env var. Zero code change.

Start free → Read the docs →

Free tier · BYOK · MIT self-host · Cancel anytime

-ANTHROPIC_BASE_URL=https://api.anthropic.com +ANTHROPIC_BASE_URL=https://api.prxy.monster +PRXY_KEY=prxy_xxx # your Anthropic key still does the inference

Modules run before every call · optimize, cache, remember, and cap spend

Built for what broke this month

Real incidents → real modules.

What broke Where The module
Auto-compaction regression dropping user intent mid-session Issue #36068 · Mar 19, 2026 Compaction Bridge
MCP tool definitions burning 67K–143K tokens before you type Apideck post · Mar 17, 2026 MCP Optimizer
Uber's $3.4B AI budget exhausted by April CTO disclosure · Apr 15, 2026 Cost Guard
Claude Code v2.1.89 → 3–50× faster rate-limit drain March 2026 release Semantic + Exact Cache
Quiet Max-only pricing test on Claude Code Apr 22, 2026 MIT self-host
Context rot after ~2 hours of session Widely reported · Apr 2026 IPC + Rehydrator

What you actually get

It's a proxy.
That's it.

You send a request to api.prxy.monster with your existing Anthropic, OpenAI, or Bedrock key. The request flows through your configured module pipeline — caching, MCP optimization, pattern injection, cost guards — then hits your provider with your key. The response comes back the same way. Same wire format you already use.

1 · You send

curl -X POST https://api.prxy.monster/v1/messages \
  -H "Authorization: Bearer $PRXY_KEY" \
  -H "X-Provider-Key: $ANTHROPIC_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "claude-sonnet-4-6",
    "max_tokens": 200,
    "messages": [{"role":"user","content":"Hello, prxy."}]
  }'

SDK drop-in: just swap ANTHROPIC_BASE_URL

2 · The pipeline runs

→ mcp-optimizer       # prune tool defs to what this request needs
→ semantic-cache      # similar requests return cached
→ patterns            # inject relevant past solutions
→ cost-guard          # short-circuit if budget breached
→ your provider       # using your key, billed to your account

Toggle modules per key via PRXY_PIPE

3 · You get back

Standard Anthropic / OpenAI response shape. SDKs work unchanged. Usage attribution per-request, scoped to your account.

Cache hit? Returns instantly with zero provider tokens consumed.

~23% of real-workload calls return cached

prxy.monster does not bill you for tokens. Your provider bills you for tokens. We bill you for the gateway and the module pipeline. We never mark up inference.

Not an inference provider. Not a web proxy. Not a VPN. Not prxy.com.

How it works

01
Point & forget

Replace one env var. Zero code changes. Every app, every framework, every model — it just works.

02
It learns

Every conversation forges patterns. Outcomes are tracked. Failures retire. Good solutions reinforce.

03
Compounds

Patterns inject before each request. Context never resets. Your AI bill goes down over time.

Featured modules

Each module fixes a specific thing that broke.

See all 12 modules →

vs the field

Other proxies route.
prxy.monster remembers.

prxy.monster OpenRouter Portkey Helicone LiteLLM
Touches your provider bill NEVER YES — markup YES — resold YES — proxied N/A
Per-request pricing (not per-token) YES NO NO NO N/A
Free tier with all base modules YES — 1k req/mo NO NO limited YES — self-host
Multi-provider routing (BYOK) YES YES YES YES YES
MCP token optimization YES NO NO NO NO
Infinite context (compressed) YES NO NO NO NO
Pattern learning across sessions YES NO NO NO NO
Semantic cache YES NO YES NO NO
Self-host (MIT/Apache) YES — MIT NO YES — Apache 2.0 YES YES
Composable modules YES NO NO NO NO

Most gateways are routers. prxy.monster is the modules.

Plays nice

Works with what you already use.

Same wire format as Anthropic and OpenAI. Most integrations are a single env var. Zero code change for the SDKs you already have wired up.

AI Coding Tools

30 sec Claude Code One env var. Kills the 45-min rate-limit cliff. 2 min OpenClaw Copy-paste JSON5 provider config. 1 setting Cursor Custom OpenAI Base URL field. 2 fields Cline OpenAI Compatible provider settings. 2 flags Aider --openai-api-base + --openai-api-key. 1 line Continue.dev apiBase in ~/.continue/config.json.

SDKs

1 env var Anthropic SDK ANTHROPIC_BASE_URL. Zero code change. 1 env var OpenAI SDK OPENAI_BASE_URL. Zero code change. 1 env var Vercel AI SDK @ai-sdk/openai + @ai-sdk/anthropic. 1 line Google GenAI @google/genai via OpenAI-compat.

Frameworks

1 arg LangChain JS configuration.baseURL on ChatOpenAI. 1 arg LangChain Python base_url on ChatOpenAI / ChatAnthropic. 1 arg LlamaIndex api_base / additionalSessionOptions.baseURL. 1 env var Mastra Inherits from the AI SDK provider config. 1 line Instructor Wraps OpenAI/Anthropic — same swap. 0 changes AI Elements UI components inherit from AI SDK config.

All integration guides →

Deploy

Cloud or local. Same modules.

Cloud
api.prxy.monster

Hosted gateway. Zero ops. Account-scoped memory and cache.

  • API keys, usage, billing, and team workflows managed for you
  • Patterns + cache available across your workspace
  • Bring your own provider key or use configured provider routes
  • Cost guards and hard limits available per key
Local
prxy-local

Single local gateway. Private data volume. MIT licensed.

  • Runs on your laptop, your homelab, your VPS
  • No telemetry. Patterns stay on your machine.
  • Same composable modules as cloud
  • BYOK to whichever providers you trust
Self-deploy
Private cloud

Dedicated deployment for teams that need their own account boundary.

  • Your data stays inside your controlled environment
  • Dedicated provider routing and policy controls
  • Custom domain, workspace policy, and support path
  • Best for regulated teams and sensitive codebases

What it costs
to run your stuff
through the monster.

Requests, not tokens. Your provider already charges you per token — we don't double-dip.

prxy_FREE

$0 forever

1,000 requests / month

  • Bring your own provider keys
  • All base modules (cache, mcp-optimizer, cost-guard)
  • Account-scoped patterns
  • Self-host unlimited — MIT-licensed
  • Community Discord
Start free →

prxy_PRO

$20 / month

100,000 requests / month · then $0.20 per 1k

  • Everything in Free
  • Pattern memory shared across your projects
  • Custom pipeline configs per API key
  • MPP merchant endpoint (agents pay you per call)
  • Email support
Go Pro →

prxy_TEAM

$99 / month

1,000,000 requests / month · then $0.10 per 1k

  • Everything in Pro
  • Custom modules (your code in our pipeline)
  • Per-seat attribution + spend caps
  • SSO + audit log
  • Priority support
Go Team →

One request = one HTTP call into our gateway. Streaming counts as one. Cached hits count as one. Failed-upstream calls don't count. Your provider bill (Anthropic, OpenAI, Bedrock) is paid directly to them at their list rates — we never see it.

Try it
in 30 seconds.

Subscribe to Pro, get an API key by email, then paste this in your terminal. Same wire format as Anthropic, so any SDK works.

curl -X POST https://api.prxy.monster/v1/messages \ -H "Authorization: Bearer $PRXY_KEY" \ -H "Content-Type: application/json" \ -d '{ "model": "claude-sonnet-4-6", "max_tokens": 200, "messages": [{"role":"user","content":"Hello, prxy."}] }' # → Your request runs through the active module pipeline # → Swap model names to any provider route configured for your account

Common questions

Frequently asked.

Is this an Anthropic SDK drop-in?
Yes. Set ANTHROPIC_BASE_URL=https://api.prxy.monster and provide your Anthropic key as a header. Existing code keeps working. OpenAI SDK works the same way via OPENAI_BASE_URL.
Whose key pays for the tokens?
Yours. prxy.monster is BYOK — every request hits your provider on your account. We never touch your provider bill and never mark up inference. We charge for the module pipeline only.
What's the module pipeline do?
Every request flows through configurable middleware: mcp-optimizer, semantic-cache, exact-cache, patterns, cost-guard, compaction-bridge, ipc, rehydrator, prompt-optimizer, tool-cache, router, guardrails. You toggle modules per API key via PRXY_PIPE. Same primitives across cloud + self-hosted.
How does pricing work?
Free tier covers 1,000 requests/month. Pro is $20/mo with 100K requests, then $0.20 per 1,000 overage. Team is $99/mo with 1M requests, then $0.10 per 1,000 overage. Provider tokens are billed by your provider account; prxy.monster does not mark them up.
Where does my data go?
Cloud: requests pass through the managed gateway, and patterns/caches are scoped to your account. Local: prxy-local runs on your machine with no telemetry; data lives in your local volume.
Open source?
prxy-monster-local, prxy-module-sdk, and prxy-cli are MIT-licensed on npm (npm i -g prxy-cli). Hosted control plane and any future paid modules are closed source.
Can agents pay per-call?
Agent-payment discovery is in private preview. API-key access is the production path today.
Cancel anytime?
Yes. Stripe customer portal, one click. No retention games. Your API key keeps working until the end of the billing period.

Start
today.

Create your account, choose a plan, and continue through Stripe Checkout. When payment succeeds, your prxy_ API key is provisioned and emailed automatically.

No password required at checkout. Account is provisioned when payment succeeds. We email a magic-link to sign in at lair.prxy.monster.

Already have an account? Sign in →

✓ Opening Stripe Checkout...
Something went wrong. Try again or email [email protected].