AI Operations Intelligence

Stop flying blind
on your AI costs

Real-time LLM pricing, model benchmarks, and deprecation alerts — delivered via MCP-native API. For humans and AI agents.

Start for $99/mo View endpoints →
$ curl https://agentlayer-api.agentlayer.workers.dev/v1/llm-costs \
  -H "x-api-key: al_live_..."
 
{ "data": [
  { "provider": "Anthropic", "model_id": "claude-sonnet-4-6",
    "input_price_per_1m_tokens": 3.00, "output": 15.00 },
  { "provider": "OpenAI", "model_id": "gpt-4o",
    "input_price_per_1m_tokens": 2.50, "output": 10.00 }
], "count": 8 }
8
LLM providers tracked
5
Intelligence endpoints
<1h
Price update latency
30d
Deprecation advance notice
API Reference

Everything your AI stack
needs to know

Five production-ready endpoints. MCP-native. Auth via API key. JSON responses. No SDKs needed.

GET
/v1/llm-costs
Real-time token pricing across OpenAI, Anthropic, Google, and Mistral. Input and output prices per 1M tokens with context window sizes.
Live
GET
/v1/model-benchmarks
MMLU, HumanEval, and MATH scores per model. Compare performance across providers before committing to an architecture decision.
Live
GET
/v1/api-deprecations
Upcoming model deprecations with end-of-life dates and migration targets. Never get caught in an emergency migration again.
Live
GET
/v1/quality-scores
Aggregated community quality signals from Reddit, HN, and HuggingFace. Know which models are actually performing in the wild.
Live
GET
/v1/agent-spend
Industry benchmark data on typical LLM spend by company size and agent type. Know if you're paying too much before your CFO asks.
Coming Soon
Live Data

Current LLM costs,
pulled right now

Fetched live from /v1/llm-costs. Refreshed every 6 hours via automated scrape.

GET /v1/llm-costs Loading...
Fetching live pricing data...
Use Cases

Built for teams
shipping with AI

Whether you're a solo developer or running a fleet of AI agents, AgentLayer keeps your stack informed.

🤖
AI Agents
Your agents can query AgentLayer directly via MCP to make cost-aware model selection decisions autonomously. No human in the loop required.
⚙️
Platform Engineers
Wire AgentLayer into your observability stack. Get webhook alerts when prices move or deprecations are announced. Sleep easier.
📊
AI Product Teams
Make model decisions with data instead of vibes. Compare benchmarks, track cost trends, and justify every infrastructure choice.
Pricing

Simple pricing,
serious intelligence

All plans include all live endpoints. Upgrade for higher rate limits and priority support.

Starter
$99
per month
  • All 4 live endpoints
  • 100 requests / minute
  • JSON API access
  • Email support
  • API key provisioned instantly
Get started
Most popular
Pro
$299
per month
  • All 4 live endpoints + agent-spend (Q2)
  • 1,000 requests / minute
  • Webhook alerts on price changes
  • Priority support (2h response)
  • API key provisioned instantly
Get started
Scale
$499
per month
  • All 5 endpoints
  • 5,000 requests / minute
  • Webhook alerts + Slack integration
  • Agent spend benchmarks
  • Dedicated support
Get started
Get started today

Your AI stack is spending
more than it needs to

Join teams using AgentLayer to keep their LLM costs under control.