Stop flying blind
on your AI costs & compute.
Real-time LLM pricing, benchmarks, and deprecation alerts — plus sovereign compute with zero-trust isolation, transparent metering, and SOC2-ready audit trails.
✅ No credit card required • Instant API key • Cancel anytime
tracked
endpoints
latency
notice
See live LLM pricing in action
Click "Run" to fetch real-time pricing data from our API. No account needed.
/api/public/llm-costs
Click "Run" to see live pricing data...
This is the same data our MCP server and API provide. Get an API key for full access.
Live LLM Pricing
Real-time cost data directly from our intelligence engine. Search, filter, and compare top models instantly.
| Model ⇅ | Provider ⇅ | Input Cost / 1M ⇅ | Output Cost / 1M ⇅ | Context Window ⇅ |
|---|---|---|---|---|
| Fetching live pricing... | ||||
Calculate Your LLM Savings
See how much you could save by intelligently routing requests across models based on task complexity.
curl -H "x-api-key: YOUR_API_KEY" \
https://apipoints-worker.francis-e3b.workers.dev/v1/llm-costsOperational Intelligence for AI Systems
Real-time LLM pricing, cross-provider comparisons, and model-level benchmark data delivered via MCP-native API endpoints.
Live LLM Pricing
Track 8+ LLM providers with normalised cost per 1M tokens. Input/output pricing, model IDs and versions. Hourly updates.
Intelligence Endpoints
Access /v1/llm-costs, /v1/model-benchmarks, /v1/deprecations, /v1/providers, /v1/changes via REST API.
Deprecation Alerts
30-day advance notice for provider and model-level deprecations. Delivered via API for proactive management.
MCP-Native Integration
Built for Claude MCP, Cursor MCP, Windsurf MCP. Any MCP-enabled agent can access intelligence endpoints.
Real-Time Updates
Hourly price updates across all providers. Stay current with the latest LLM pricing changes as they happen.
API Key System
Secure key generation with usage tracking. Tiered access levels for Starter, Pro, and Scale plans.
Model Benchmarks
Access performance benchmarks across models. Compare quality, speed, and cost metrics in one place.
Provider Change Tracking
Monitor pricing changes, new model releases, and provider updates. Historical data for cost analysis.
How APIPoints Works
Simple API access to real-time AI operations intelligence. No infrastructure to manage.
Get API Key
Sign up and generate your API key. Starter plan includes full access to all intelligence endpoints.
Query Intelligence
Access real-time LLM pricing, model benchmarks, and deprecation alerts via REST API or MCP.
Integrate via MCP
Connect Claude, Cursor, or Windsurf via MCP-native endpoints. Get intelligence directly in your workflow.
Hourly updates • 8+ providers • MCP-native
Intelligence Endpoints
MCP-native API endpoints delivering real-time LLM pricing, benchmarks, and alerts.
/v1/llm-costs
Real-time LLM pricing across all providers
/v1/model-benchmarks
Performance benchmarks comparing models
/v1/deprecations
Active and upcoming deprecations
/v1/providers
List all tracked LLM providers
/v1/changes
Recent pricing changes and updates
MCP-native • All endpoints support x-api-key auth
APIPOINTS Compute
Sovereign execution layer for AI agent workloads. Enterprise identity, rate limits, usage metering, spend caps, and audit logging — powered by Daytona.
Zero-Trust Isolation
Dedicated kernel sandboxes with hardware-level isolation. No shared tenants.
Data Sovereignty
Regional data residency with egress network filtering and IP whitelisting.
SOC2 Aligned
SHA-256 hashed audit logs with immutable ledger entries for compliance.
Sub-100ms Cold Start
Warm-pool sandbox provisioning. GPU nodes ready in milliseconds.
Sovereign Execution Pipeline
APIPOINTS Gateway
mTLS • JWT Auth • Rate Limiting • API Key Verification
Governance Engine
Egress Filter • Spend Cap Check • Concurrency Limits
Isolated Execution
Daytona MicroVM • GPU Node • Code Sandbox
Metering & Ledger
Real-time Billing • SHA-256 Hashed Logs • SOC2 Audit
Raw Compute vs. APIPOINTS Managed
Quickstart
pip install apipoints-compute
from apipoints_compute import APIPointsCompute
client = APIPointsCompute("apk_live_...")
sbx = client.sandboxes.create(
vcpu_count=2,
memory_mb=1024
)
result = sbx.run_code(
"print('Hello from APIPOINTS')"
)
npm install @apipoints/compute
import { APIPointsCompute } from
"@apipoints/compute";
const client = new APIPointsCompute(
"apk_live_..."
);
const sbx = await client.createSandbox({
vcpu_count: 2,
memory_mb: 1024
});
const result = await sbx.runCode(
"console.log('Hello from APIPOINTS')"
);
Compute Endpoints
/v1/compute/sandboxesCreate isolated sandbox
/v1/compute/sandboxes/:id/code-runExecute code in sandbox
/v1/compute/snapshotsCreate persistent snapshot
/v1/compute/desktopsProvision VNC/WebRTC desktop
/v1/compute/sandboxes/:id/logs/streamSSE log streaming
/v1/compute/usageUsage summary & balance
Calculate Your Agent Margin
Transparent cost breakdowns. See exactly what you pay vs. raw wholesale compute.
Simple, Transparent Pricing
Start free. Upgrade when you need more. All plans include MCP-native access.
Explore the API and test basic intelligence endpoints.
- 5,000 API credits/month
- 10 requests/min rate limit
- All 5 data endpoints
- Community support
- No compute sandbox access
Full intelligence access with $49/mo compute credits — 100% rollover.
- All 5 intelligence endpoints
- 60 requests/min rate limit
- $49 compute credits/mo
- 3 concurrent sandboxes
- vCPU $0.0655/hr
- 1 seat included
Advanced intelligence + $149 compute credits. GPU sandbox access.
- All 5 intelligence endpoints
- 300 requests/min rate limit
- $149 compute credits/mo
- 15 concurrent sandboxes
- GPU access (RTX 4090 $1.24/hr)
- 5 team seats included
Maximum compute + intelligence. H100 access, custom SLAs.
- All 5 intelligence endpoints
- 1,000 requests/min rate limit
- $499 compute credits/mo
- 50 concurrent sandboxes
- H100 access ($4.55/hr)
- 20 team seats + dedicated support
Documentation
Everything you need to start using APIPoints in under 5 minutes.
Get Your API Key
Choose a plan and get instant access to your API key. Your key looks like al_live_... and is valid for 30 days.
Make Your First API Call
Use curl or any HTTP client. All requests need your API key in the x-api-key header.
curl https://apipoints-worker.francis-e3b.workers.dev/v1/llm-costs \ -H "x-api-key: YOUR_API_KEY"
Available Endpoints (5 Total)
/v1/llm-costs
Real-time LLM pricing
/v1/model-benchmarks
Model performance data
/v1/deprecations
30-day advance notices
/v1/providers
All tracked providers
/v1/changes
Recent pricing changes
Connect via MCP (Optional)
APIPoints is MCP-native. Connect directly to Claude, Cursor, or Windsurf for real-time intelligence in your AI workflow.
// Claude MCP Configuration
{
"mcpServers": {
"APIPoints": {
"command": "npx",
"args": ["-y", "@apipoints/mcp-server"],
"env": {
"APIPOINTS_API_KEY": "YOUR_API_KEY"
}
}
}
}