Lynkr LLM router outperforms GPT-5 on academic benchmark

Lynkr, an LLM router, was evaluated on the academic RouterArena benchmark, which assesses performance across various metrics including accuracy, cost, and robustness. The results placed Lynkr mid-table among 27 routers, but it notably outperformed GPT-5 when used as a router, achieving a higher Arena score at a significantly lower cost. Lynkr also demonstrated strong robustness, with its routing decisions remaining stable even when queries were rephrased. AI

IMPACT Demonstrates that specialized routing systems can outperform frontier models for cost-efficiency and robustness in query routing.

RANK_REASON Academic benchmark evaluation of an LLM router. [lever_c_demoted from research: ic=1 ai=1.0]

Read on dev.to — LLM tag →

• GPT-5
• OpenRouter
• RouterArena
• RouteWorks

• product
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →