One API.
Every model.

Access leading AI models through a single, fast and reliable API. One endpoint, zero configuration.

curl https://vrouter.my.id/v1/chat/completions \
  -H "Authorization: Bearer *** \
  -H "Content-Type: application/json" \
  -d '{
    "model": "nemotron-3.5-lite",
    "messages": [
      {"role": "user", "content": "Hello!"}
    ]
  }'
API ready
https://vrouter.my.id/v1

All major providers, one endpoint

Access models from leading AI labs through a unified API. No per-provider setup required.

OpenAI
Active
Anthropic
Active
Google
Active
DeepSeek
Active
Meta
Active
Mistral
Active
Cohere
Active
Qwen
Active

Built for production

Infrastructure that scales with your application. Every feature designed for reliability.

Unified API

Single endpoint for every model. Change the model name in your request, not your code.

Multiple Providers

OpenAI, Anthropic, Google, DeepSeek, Meta, Mistral, and more — all through one key.

OpenAI Compatible

Drop-in replacement. Change your base URL and every existing SDK works unchanged.

Streaming

Full SSE streaming support. Token-by-token responses for real-time UIs.

Fast Routing

Sub-50ms routing latency. Automatic failover and load balancing across providers.

Usage Monitoring

Real-time dashboards for token usage, costs, and per-model latency breakdowns.

Up and running in 3 steps

No configuration. No SDK changes. Just set a URL and start building.

1
Sign up

Create your account

Free tier included. No credit card required. Get your API key in seconds.

# One command to verify
curl https://vrouter.my.id/v1/models \
  -H "Authorization: Bearer ***
2
Get API key

Set your base URL

Replace the OpenAI base URL with VRouter LLM API. Existing SDKs work unchanged.

# Python
client = OpenAI(
  base_url="https://vrouter.my.id/v1",
  api_key="YOUR_KEY"
)
3
Start building

Make your first request

Send a completion request. Switch models by changing the model field.

# Switch models instantly
response = client.chat.completions.create(
  model="nemotron-3.5-lite", # or "qwen3.7-flash", "deepseek-v4-flash"...
  messages=[{"role": "user", "content": "Hello!"}]
)

Simple, transparent pricing

Start free. Scale without limits. No hidden fees.

Free

For hobbyists and testing

Free /mo
  • ✓ 500 requests / month
  • ✓ Access to core models
  • ✓ OpenAI-compatible API
  • ✓ Streaming support
  • — Load balancing
  • — Custom provider keys
Start Free
Most Popular

Pro

For teams and production

$20 /mo
  • ✓ 5,000 requests / month
  • ✓ All models + gated access
  • ✓ Auto load balancing
  • ✓ Custom provider keys
  • ✓ Priority support
  • ✓ Usage analytics
Upgrade to Pro

Enterprise

For high-volume production

Custom
  • ✓ Unlimited usage
  • ✓ Dedicated instances
  • ✓ Custom SLAs
  • ✓ SSO / SAML
  • ✓ 24/7 priority support
  • ✓ Volume discounts
Contact Sales

Compare plans

Feature Free Pro Enterprise
Requests / month 500 5,000 Unlimited
Models available Core All + gated All + custom
Rate limit Standard High Custom
Streaming ✓ ✓ ✓
Load balancing — ✓ ✓
Custom provider keys — ✓ ✓
Support level Community Priority 24/7
Usage analytics — ✓ ✓

[TO CONFIRM] — Verify exact plan limits before launch.

Frequently asked questions

Everything you need before making your first request.

500 requests per month with access to core models. No credit card required. Sign up and get your API key in seconds.
236+ models from providers including OpenAI, Anthropic, Google, Meta, Mistral, DeepSeek, Qwen, and more. Free tier models are unlimited. Pro tier unlocks the full catalog including gated models.
Yes. VRouter LLM API is fully OpenAI-compatible. Any tool that supports OpenAI's Chat Completions API works — Cursor, Claude Code, OpenAI Codex, Continue.dev, Aider, Windsurf, and more. Just set your base URL to https://vrouter.my.id/v1.
No. We only store usage metadata (model name, token counts, latency). Your prompts and completions are never logged, stored, or shared with third parties.
Use your API key as a Bearer token: Authorization: Bearer *** The API is fully OpenAI-compatible, so existing SDKs work unchanged.
Your access falls back to the free tier models. You can continue using core models without interruption. No downtime, no errors.

Start building today

Free tier. No credit card. One API key, every model.

Quick start 30 seconds
# Install the OpenAI SDK
pip install openai

# Set your base URL and key
client = OpenAI(
  base_url="https://vrouter.my.id/v1",
  api_key="YOUR_KEY"
)

# Start chatting with any model
response = client.chat.completions.create(
  model="nemotron-3.5-lite",
  messages=[{"role": "user", "content": "Hello!"}]
)