OneRouter

Infron AI is the world's first AI model marketplace and inference provider routing platform, offering developers a unified API interface to access over 300 open-source and commercial AI models. As an AI infrastructure layer, Infron achieves cross-provider high availability, seamless developer workflows, and ultra-low-cost scaling through the Infron routing stack.

The platform supports various AI capabilities including text generation, image generation, video generation, audio generation, search, embeddings, and batch processing. Compatible with OpenAI SDK and Anthropic SDK, developers can switch models without modifying code. Infron's core advantages lie in intelligent routing, automatic failover, cost optimization, and zero data retention policies, ensuring enterprises use AI efficiently and securely.

The platform offers two pricing models: pay-as-you-go and enterprise custom plans. Through unified billing, automatic capacity management, and SLA guarantees, it helps enterprises save up to 35% on AI spending. Infron serves companies including YC W26-incubated Pax Historia, YTL AI Labs, and Agnes AI, processing over 6 trillion tokens monthly while maintaining 99.9% uptime.

• Unified API Access to 300+ Models: Access mainstream AI models like GPT-5.3, Claude Opus 4.6, Gemini 3 Flash, Llama 4, Qwen 3.5 through a single API, supporting text, image, video, audio, search, embeddings, and batch processing modalities
• Intelligent Routing and Automatic Failover: Automatically selects the best model provider based on performance, cost, and availability, switching to backup providers when failures occur, ensuring 99.9% service availability
• Cost Optimization: Helps enterprises save up to 35% on AI spending through intelligent routing and bulk discounts, with platform fees of only 5%+$0.35 (pay-as-you-go) or 3% (enterprise), model prices match original providers without markup
• Zero Data Retention (ZDR): Does not log user prompts and responses by default, only stores basic request metadata (timestamps, models used, token counts), ensuring data privacy and compliance
• Prompt Caching: Automatically enables prompt caching to reduce token consumption for repeated requests, lowering latency and saving costs
• Structured Outputs: Supports JSON Schema-defined structured output formats, ensuring AI responses conform to expected data structures
• Tool Calling: Supports function calling and tool integration, allowing AI models to call external APIs and tools for complex tasks
• Multimodal Input: Supports mixing text, images, PDFs, and audio in the same request, models can process multiple data types simultaneously
• Reasoning and Thinking Modes: Supports thinking modes for models like Gemini 3 and Claude Opus, improving accuracy on complex tasks through multi-step reasoning
• 1M Token Long Context Window: Supports ultra-long context processing for document analysis, code review, and long conversation scenarios
• Batch Processing API: Supports batch request processing for large-scale data processing and offline tasks
• Performance Monitoring and Analysis: Provides real-time usage metrics, cost tracking, and performance analysis to help developers optimize AI applications
• Billing Transparency: Detailed billing logs and usage statistics, supports filtering by API key, model, and provider
• BYOK (Bring Your Own Key): Supports users accessing model providers with their own API keys, 0% platform fee
• SDK and Framework Integration: Compatible with OpenAI SDK, Anthropic SDK, LangChain, PydanticAI, Langfuse, n8n, LiteLLM, and other mainstream frameworks and tools
• Enterprise-Grade Security and Compliance: Fine-grained data policies, policy-based routing, managed policy enforcement, and SLA-guaranteed uptime

• Step 1: Register Account: Visit https://infron.ai/login, register with email or sign in with Google account
• Step 2: Create API Key: After logging in, go to "API Key" page, click "Create api key" button to generate API key, copy and securely store the key
• Step 3: Set Up Billing: Go to "Billing" page, click "Add Payment Method" to add payment method, enable "Low Balance Alert" to receive email reminders when balance is low
• Step 4: Top Up Account: Add credits on Credits page, platform uses prepaid model, fees are deducted from balance when using API
• Step 5: Select Model: Visit Model Marketplace to browse 300+ available models, select appropriate model based on task requirements
• Step 6: Integrate API: Use OpenAI SDK or Anthropic SDK, point API endpoint to Infron (e.g., https://api.infron.ai/v1), authenticate with generated API key
• Step 7: Make Requests: Send requests according to OpenAI or Anthropic API specifications, Infron automatically routes to best provider and returns response
• Step 8: Monitor Usage: View real-time usage metrics, cost statistics, and performance analysis on Activity page, view detailed request logs on Logs page
• Step 9: Optimize Configuration: Adjust routing strategies, enable failover, and configure budget controls based on usage to optimize cost and performance
• Step 10: Scale Application: As business grows, upgrade to enterprise version for lower platform fees, dedicated support, and SLA guarantees

• Single Integration, Access All Models: Access 300+ models through one API without separate integration for each provider, significantly reducing development and maintenance costs
• High Availability Guarantee: Distributed infrastructure and automatic failover mechanisms ensure 99.9% uptime, automatically switching to backup providers when failures occur
• Significant Cost Reduction: Save up to 35% on AI spending through intelligent routing and bulk discounts, enterprise users enjoy lower platform fees and model discounts
• Edge Deployment Low Latency: Infron runs at the edge, minimizing latency between users and inference services for faster response times
• Enterprise-Grade Security and Compliance: Zero data retention policy, fine-grained data policies, and policy-based routing ensure prompts only go to trusted models and providers
• Developer Friendly: Compatible with OpenAI SDK and Anthropic SDK, switch models without code changes, supports mainstream frameworks like LangChain and PydanticAI
• Unified Billing and Management: Single contract, unified billing, and centralized management, avoiding complexity of managing 30+ provider relationships
• Automatic Capacity Management: Automatically uses Infron's capacity pool when users hit rate limits, ensuring uninterrupted service
• Real-time Monitoring and Analysis: Provides detailed usage metrics, cost tracking, and performance analysis to help developers optimize AI applications
• Flexible Payment Methods: Supports credit cards, Alipay, PayPal, WeChat Pay, and invoice payment, enterprise users can apply for credit lines
• Rapid Deployment: Integrate with just a few lines of code, no code rewrite needed from prototype to production, accelerating product launch

Tier	Price	Description
Pay-as-you-go	5% + $0.35/transaction	For individual developers and small teams, model prices match original providers without markup, supports credit cards, Alipay, PayPal, etc.
Enterprise	3%/transaction	For enterprise users, model prices up to 30% discount, provides SLA guarantee, 24/7 support, dedicated Slack or WhatsApp support channels, data protection agreement, and invoice payment
BYOK (Bring Your Own Key)	0% platform fee	Users access model providers with their own API keys, Infron charges no platform fee, suitable for users with existing provider accounts

Fee Details:

• Platform fees cover credit card transaction fees, currency conversion fees, taxes, and other costs
• Model pricing details available at Model Marketplace, each model billed per million tokens, typically different prices for prompt and response tokens
• Enterprise users enjoy tiered discounts based on actual monthly usage, up to 30% savings
• Top-up example: Top up $50, platform fee $2.85 (5%+$0.35), actual credit $50
• Enterprise users can get full refund for unused account balance with no processing fees
• Pay-as-you-go user fees are non-refundable

• Official Documentation: Visit https://infron.ai/docs for complete API documentation, quick start guides, and integration examples
• Email Support: Send emails to support@infron.ai or info@infron.ai for technical support
• Book Demo: Visit https://infron.ai/contact to book product demo or consult enterprise plans
• Community Support: Join Discord community or follow X (Twitter) account for latest updates and community help
• Enterprise Support: Enterprise users get 24/7 priority support through dedicated Slack or WhatsApp support channels
• Service Status: Visit https://status.infron.ai/ to check real-time service status and historical uptime
• Blog Resources: Visit https://infron.ai/blog to read technical articles, best practices, and product updates
• Privacy Policy: View Privacy Policy to understand data processing and privacy protection measures
• Terms of Service: View Terms of Service to understand usage rules and liability terms

Infron AI is a web-based API service platform that primarily provides services through API interfaces, requiring no client downloads. Developers can directly integrate Infron API into applications or access through the following methods:

• Web Console: Visit https://app.onerouter.pro/login to log into web console, manage API keys, view usage, and configure account
• OpenAI SDK Integration: Use OpenAI SDK in Python, JavaScript, and other languages, point API endpoint to Infron
• Anthropic SDK Integration: When using Anthropic SDK, point API endpoint to Infron for seamless switching
• LangChain Integration: Configure Infron as LLM provider in LangChain projects
• PydanticAI Integration: Use Infron API in PydanticAI projects
• n8n Integration: Add Infron nodes in n8n workflows
• LiteLLM Integration: Use Infron through LiteLLM proxy
• Command Line Tools: Use OpenAI Codex CLI or OpenCode and other command line tools, configure Infron API endpoint
• Browser Extensions: Some browser extensions support custom API endpoint configuration, can be configured to use Infron

For specific integration methods, refer to the framework and integration guides in Official Documentation.