Layerr is a 12-layer orchestration intelligence system that sits between your applications and LLM providers. A dedicated router model classifies every prompt into 8 intent categories, then a workload analyzer, workspace selector, 12-dimensional scoring engine, and strategy engine determine the optimal model. Fallback chains, provider health checks, explainability, replay intelligence, adaptive learning, and benchmark harnesses are all built in. Currently v0.8.0-alpha - early but real.

Does Layerr generate AI responses?

No. Layerr never generates responses itself. It acts as an intelligent routing and orchestration layer, forwarding your prompts to existing providers like Anthropic, OpenAI, Mistral, Google, and others, then returning the response back to your tool.

Is it compatible with Claude Code and Cursor?

Yes. Layerr exposes a fully OpenAI-compatible API endpoint. Set OPENAI_BASE_URL to Layerr and Claude Code, Cursor, Roo Code, or any OpenAI SDK client routes through it instantly. Zero code changes. The author has been running this against Claude Code for weeks - local models for quick patches, escalated to sonnet for heavy reasoning.

How does intelligent routing work?

Layer 1: a router model classifies intent into CODING, REASONING, CREATIVE, VISION, DOCUMENT, FAST, SECURITY, or GENERAL. Layer 2: deterministic workload analysis scores complexity, latency sensitivity, cost sensitivity, context size, and reasoning depth with zero LLM calls. Layer 3: a workspace profile (General, Coding, Research, Creative, Local-First, Cost, or Performance) is selected. Layer 4: the strategy engine builds a 12-dimensional scoring profile with latency tolerance, cost sensitivity, reasoning depth, and fallback thresholds. Layer 5: every candidate model is scored across 12 dimensions plus runtime enrichment. Layer 6: the primary model is selected and a fallback chain of up to 4 candidates is prepared. If the primary fails, the next candidate is tried automatically. Every decision produces a structured explainability output.

Can I self-host Layerr?

Yes. The Layerr runtime deploys via Docker or Kubernetes on any cloud or bare metal. Self-hosted deployments use your own API keys, run all routing logic locally, and receive all 9 capabilities including workspace customization, replay intelligence, benchmark harness, and entitlements. In v0.8.0-alpha the focus is on getting the local + cloud hybrid right first.

What models are supported?

Any model available through OpenRouter (100+ models), plus direct integrations with Anthropic, OpenAI, Mistral, Google Gemini, Cohere, and Ollama for local inference. The router is currently qwen2.5:14b or similar. If a provider exposes an API, Layerr can route to it.

Is my code or data sent to Layerr?

Only routing metadata is processed - estimated prompt complexity, context size, and task type. Your actual code and prompts go directly to the chosen LLM provider. Self-hosted deployments send zero data to Layerr servers. The router model runs locally if you self-host.

v0.8.0-alpha - early but real. Install it, try it, tell us what breaks.

Same agents. Same workflow.
Smaller bill.

Layerr is a self-hosted orchestration runtime. It reads every prompt, classifies what it needs, and picks the optimal model; local, cheap cloud, or premium. OpenAI-compatible. Your keys. Your machine. Every routing decision explained.

View Docs

OpenAI CompatibleOpenRouter ReadySelf-HostedBYOKDocker Deployable

layerr-engine · orchestration tracereq_9x8k2mrouting

AI Workloads

Chatbot

conversational

RAG Pipeline

retrieval + gen

Agentic Flow

multi-step tools

prompt

Engine

Layerr Engine

complexitycostcontextrisk

route

Model Routing

gpt-4o-mini

attempt 1

conf: 42%SKIP

claude-3-haiku

attempt 2

conf: 71%SKIP

claude-sonnet-4

escalation

conf: 89%SKIP

total latency: 250mscost: $0.089

How It Works

From Prompt to Provider in 5 Steps

Intent classification, workspace selection, 12-dimensional scoring, fallback chains, and full explainability - every time.

Client Sends Prompt

Your application sends a request to the Layerr endpoint. No SDK changes required. The router model immediately classifies intent into 8 categories.

Workload Analysis

Deterministic analysis of complexity, latency sensitivity, cost sensitivity, context size, and reasoning depth - zero LLM calls. Then picks the right workspace profile.

Strategy Engine Scores

12-dimensional scoring profiles evaluate every candidate model. Latency tolerance, cost sensitivity, reasoning depth, and fallback thresholds are all tuned by workspace.

Orchestration Executes

Primary model selected. Fallback chain ready up to 4 deep. Provider health checked. Budget enforced. If primary fails, retry with next candidate automatically.

Explained & Logged

Response returned with full explainability: routing source, confidence, top 3 candidates, policy checks, and a human-readable summary. Every decision is auditable.

intent classificationworkspace profiles12-dim scoringfallback chainsexplainabilityadaptive learning

Intelligence Layer

Not a router. An orchestration intelligence system.

Every feature was built because a solo dev got tired of babysitting models.

Intelligent Model Routing

Routes every prompt to the optimal model based on task complexity, intent, and context. Every decision in milliseconds.

Workload-Aware Orchestration

Understands what you're building - code, analysis, creative writing - and adjusts routing strategy accordingly.

Fallback Chains & Escalation

Starts cheap. Escalates when needed. No timeouts, no failures, no manual model switching.

OpenAI-Compatible API

Drop-in replacement. Set OPENAI_BASE_URL and every tool works instantly. Zero refactoring.

OpenRouter Integration

Access 100+ models with a single key. Qwen, Kimi, Mistral, and more.

Self-Hosted Runtime

Deploy on your own infra. Bring your own keys. Keep all traffic on-prem. Full data sovereignty.

Under the Hood

Layerr is not a proxy. It's an intelligent layer orchestration platform.

Workspace Profiles

Per-project configuration with model preferences, budget caps, security rules, and team defaults. Every workspace gets its own routing personality.

Strategy Engine

Define execution strategies: cheap-first, quality-first, latency-first, or balanced. The engine scores every candidate model against your strategy.

Explainability & Replay

Every routing decision includes a full reasoning trace. Re-run any past request to inspect the chain, debug decisions, or reproduce results.

Hybrid Local + Cloud

Route sensitive workloads to local models (Ollama, llama.cpp) and burst to cloud when needed. One API, both worlds.

Adaptive Learning & Benchmarks

The system learns from outcomes and improves routing accuracy over time. Built-in benchmark harness against standard datasets.

Governance & Entitlements

RBAC, rate limits, audit logs, and compliance controls. Know who used what, when, and why.

Built for

Layerr is for developers who use AI heavily and want their infrastructure to be theirs.

Solo developers running coding agents

You use Claude Code, Cursor, or Cline every day. Your monthly bill keeps creeping up. You've tried manually switching to cheaper models and it broke your flow. Layerr picks the right model for every prompt automatically, runs on your machine, and uses your keys.

Small teams sharing AI infrastructure

Your team is sharing API keys, arguing about which model to use, and watching the bill grow. Layerr gives you one self-hosted gateway that handles routing for everyone, with per-user provider mappings, shared workspaces, and budget controls.

Developers who don't trust hosted AI middleware

You don't want your prompts going through someone else's servers. You don't want vendor lock-in. You don't want surprise pricing or rate limits set by someone else. Layerr is self-hosted infrastructure that runs on your machine, with your keys, with every routing decision explained.

Integration

One Line. Every Model.

Layerr exposes a fully OpenAI-compatible endpoint. No SDK changes, no refactoring. Point your application at Layerr and every routing decision, fallback chain, and escalation happens automatically - with full explainability for every choice.

Works with any OpenAI SDK, LangChain, LlamaIndex, or custom client
Single API key manages routing across all providers
Every request traced with latency, cost, confidence, and model selected
Explainability output shows top candidates, policy checks, and reasoning

.env

# Point any OpenAI-compatible client at Layerr
OPENAI_BASE_URL=https://api.layerr.app/v1
OPENAI_API_KEY=your-layerr-key

# Every model, intelligently routed. That's it.

routing-decision.json

req_9x8k2m

{
  "request_id": "req_9x8k2m",
  "intent": "CODING",
  "intent_confidence": 0.94,
  "workspace": "coding",
  "strategy": "deep-build",
  "complexity": "medium",
  "latency_sensitive": true,
  "cost_sensitive": true,
  "selected_model": "qwen2.5-coder:14b",
  "selected_provider": "ollama",
  "selection_source": "router_llm",
  "explainability": {
    "confidence": 0.94,
    "top_candidates": [
      { "model": "qwen2.5-coder:14b", "score": 0.88 },
      { "model": "claude-sonnet-4", "score": 0.96 }
    ],
    "policy_warnings": [],
    "adaptive_reason": "local_preferred_for_coding"
  },
  "fallback_chain": [
    { "model": "qwen2.5-coder:14b", "conf": 0.88, "result": "selected" },
    { "model": "claude-3-haiku",   "conf": 0.72, "result": "skip" },
    { "model": "claude-sonnet-4",  "conf": 0.96, "result": "fallback" }
  ],
  "cost_usd": 0.0000,
  "quality_grade": "A",
  "quality_score": 92,
  "latency_ms": 300
}

Orchestration Loglive

Why Layerr

Stop Babysitting Your Models

Layerr handles the operational complexity so your team can focus on building.

Reduce AI Spend

Layerr routes every task to the cheapest model that meets the quality bar. Simple queries hit small models; complex reasoning escalates to premium. Every request is cost-optimized.

Cost-optimized

every request routed to the cheapest capable model

Stop Switching Manually

Stop managing multiple provider SDKs and API keys. One endpoint, one key, zero config. Layerr picks the right model while you focus on your product.

Zero

config changes when switching models

Context-Intelligent Routing

Layerr understands your workload before selecting a model. It considers task complexity, intent, security scope, and context size. Not just token count.

Context-aware

routing per task type and workload

Built-in Reliability

Automatic fallback chains and escalation logic mean a model outage never blocks your application. If one provider is down, Layerr silently routes to the next best option.

Resilient

automatic fallback when any provider fails

Always the Right Model

Layerr scores confidence and task complexity in real time. You always get the right model for the job, at the lowest possible cost, without ever thinking about which one to pick.

Confidence-scored

selection at every routing decision

Developer-First Infrastructure

Your Infra. Your Keys. Your Control.

The full Layerr orchestration runtime deploys on any cloud or bare metal. All traffic stays on-premises, all keys stay yours, all routing logic runs inside your perimeter.

On-premises runtime

Full control.
Zero compromise.

Deploy the entire Layerr orchestration stack inside your own perimeter. Your prompts never leave. Your keys never leave. Only licensing and intelligence updates cross the wire.

BYOK

Use your own Anthropic, OpenAI, Mistral, and Google keys

Privacy-first architecture

No prompts, code, or responses touch Layerr servers

Local orchestration engine

All routing logic runs inside your perimeter

Enterprise SSO ready

OIDC, SAML, and team-based access controls

Docker + Kubernetes

Deploy on any cloud, on-prem, or bare metal

Zero data leakage

Audit logs stay local. Full GDPR and SOC 2 alignment

Layerr Cloud

Licensing

Analytics

Intelligence

encrypted · control plane only

Your Infrastructure

Layerr Runtime

Docker / Kubernetes

running

Orchestration Engine

routing, fallback, escalation

Your API Keys

Anthropic, OpenAI, Mistral...

direct provider calls · no proxy

LLM Providers

Anthropic

OpenAI

Mistral

Google

Cohere

OpenRouter

Why I built this

I built Layerr because manually switching models in my IDE was insane, and paying premium prices for simple prompts was worse. Now it routes itself. My bill dropped. My workflow didn't change.

Aasim Pathan

founder of Layerr

Pricing

Free gets you intelligent routing. Pro gets you orchestration intelligence.

No per-seat pricing. No hidden fees. Pay for what you route.

Save 20%

All plans are self-hosted. Unlimited usage. Your keys. Your machine. Always.

Hobby

Free

Best for: Solo developers trying Layerr on personal projects

Intent classification across all configured providers
All supported LLM providers (Anthropic, OpenAI, Google, OpenRouter, Ollama, local runtimes)
Fallback chains
Community support

Popular

Pro

$99$49

Introductory offer · 50% off

Per Month

Best for: Solo developers and power users shipping with AI daily

Everything in Hobby
Execution strategies
Cheap-first, quality-first, or latency-first. Pick a strategy and Layerr applies it to every request.
Replay and trace intelligence
See exactly why each prompt was routed to a specific model. Replay any request with different settings.
Workspace profiles
Save provider mappings, model preferences, and routing rules as reusable workspaces.
Advanced analytics and benchmarking
Track cost per request, model win rates, latency distributions, and provider uptime.
Additional features being defined. Beta users help shape the tier.
Pro tier differentiation is still being finalized. Beta feedback directly influences what ships.

Coming soon

Team

$299