Welcome back to AI Router's Roundup! We're kicking off 2025 with major updates that make AI Router more accessible, private, and powerful than ever before.
We've completely revamped our pricing to make it simpler and more predictable. But that's just the beginning β we're introducing three new privacy-focused usage modes that give you unprecedented control over how you interact with LLM providers. Whether you need just the model recommendations or want full privacy for sensitive data, we've got you covered.
This edition is packed with more exciting launches: our new interactive playground for quick model testing, a comprehensive leaderboard of the best-performing models from 2024, and a special voucher for you to try all these features at no cost.
So let's jump right in.
β
Every organization has different requirements when it comes to using LLMs. Some need full control over their model infrastructure, others require maximum data privacy. We're excited to introduce three powerful modes that adapt to your specific needs:
β
βModel Routing: The Complete Solution
Our classic mode handles everything for you. We select the best model and route your requests automatically, optimizing for performance and cost. Perfect for teams who want a seamless, managed experience.
Model Selection: Stay in Control
Want to use your private models or hyperscaler infrastructure like AWS Bedrock? Our Model Selection mode gives you smart model recommendations without routing, keeping full control to call the models yourself. Perfect for teams with existing infrastructure who still want to optimize their model selection or need to follow guidelines on LLM provider selection.
Full Privacy: Maximum Security
Our most secure option: Maximum security for your sensitive data through our new embedding-based routing that identifies the optimal LLM without ever seeing your actual queries. Your sensitive data stays completely private.
NEW PRICING
Never Count Tokens Again - Just Requests
Our new simplified pricing is based on requests only, with just two straightforward plans: Free for Model Selection and 10,000 requests included and Pro for all routing capabilities.
|
|
|
β
With AI Router, we've completely avoided LLM downtime while seeing our costs steadily decrease and quality improve - all without any effort on our side.
β Florian Falk, Founder at Soji AI
|
β
Playground: Get Instant Model Recommendations
Wondering which LLM would be optimal for a specific use case? Our new Playground gives you instant model recommendations. Just paste your prompt and see which model would provide the best balance of cost, quality, and speed.
|
|
LLM Leaderboard 2024
Our analysis of over 1M+ production requests reveals which LLMs truly delivered value in 2024.
While GPT-4 dominated most of the year, Qwen 2.5 72B made a remarkable climb to challenge the top spot by year's end, showcasing how rapidly the landscape evolved with each new release.
|
Your fast track to advanced LLM routing
Try All Pro Features Free for 30 Days
Ready to explore our new privacy modes and smart fallbacks? Use code ROUNDUP to get full access to all Pro features, including 100,000 requests to test AI Router with your production workload.
|
β