Skip to main content

OpenRouter provides unified access to 200+ AI models:
  • Single API - Access models from multiple providers
  • Automatic fallbacks - Switch to similar models if unavailable
  • Competitive pricing - Transparent passthrough costs
  • No vendor lock-in - Switch models instantly
  • Model discovery - Find the best model for your task

Pricing Model

Dynamic Passthrough Pricing

OpenRouter uses passthrough pricing where you pay the vendor’s cost plus OpenRouter fee, then Splox platform fee.
Pricing Formula:Final Price = (Vendor Cost + OpenRouter Fee) + Splox Platform FeePricing updates automatically when vendors change rates.

How It Works

  1. Select a model (e.g., openai/gpt-4, anthropic/claude-3-opus)
  2. OpenRouter routes your request to the vendor
  3. You’re charged vendor cost + OpenRouter fee + Splox fee
  4. Pricing updates automatically in real-time

Example Pricing

Approximate costs (vendor + OpenRouter + Splox fees):
ProviderModelInputOutput
OpenAIGPT-4o~$3.25/M~$13.00/M
AnthropicClaude Opus 4.1~$19.50/M~$97.50/M
GoogleGemini 2.5 Pro~$1.62/M~$13.00/M
MetaLlama 3.3 70B~$0.52/M~$0.52/M
Prices are estimates and may vary. Check OpenRouter Model Directory for current vendor costs.

Available Models

OpenRouter provides access to 200+ models across categories:
Popular models across providers:
  • OpenAI: GPT-5, GPT-4o, o1, o1-mini
  • Anthropic: Claude Opus 4.1, Sonnet 4, Haiku 4
  • Google: Gemini 2.5 Pro, Gemini 2.0 Flash
  • Meta: Llama 3.3 70B, Llama 3.1 405B
  • xAI: Grok 2, Grok 3
  • Cohere: Command R+
  • Mistral: Large, Medium

Find Models

Browse the full model directory at openrouter.ai/models to see all available models, pricing, and capabilities.

Best Practices

Model Selection

Use Auto-Routing

  • Automatic fallbacks
  • Cost optimization
  • Performance balancing
  • Availability handling

Compare Models

  • Different providers
  • Various price points
  • Speed vs quality tradeoffs
  • Context window needs

Monitor Costs

  • Check current rates before high-volume use
  • Set spending limits in dashboard
  • Use cheaper models for testing
  • Monitor usage in real-time

Leverage Fallbacks

  • Primary model unavailable → fallback
  • Rate limit hit → alternative model
  • Cost threshold → cheaper option
  • Quality requirements met

Support

Need help with OpenRouter integration?