According to a recent LinkedIn post from FriendliAI, independent leaderboards from Artificial Analysis reportedly show the company’s Model APIs delivering high output speed and low latency for popular open-weight models GLM-5.1 and Gemma-4-31B. The post cites performance metrics of roughly 133 output tokens per second for GLM-5.1 and 62 tokens per second for Gemma-4-31B in non-reasoning workloads, positioning FriendliAI as faster than key third-party endpoints.
Meet Samuel – Your Personal Investing Prophet
- Start a conversation with TipRanks’ trusted, data-backed investment intelligence
- Ask Samuel about stocks, your portfolio, or the market and get instant, personalized insights in seconds
The company’s LinkedIn post highlights that these benchmarks appear to favor FriendliAI on a speed-to-latency ratio, placing its offerings in what is described as the “most attractive quadrant” for real-world inference performance. For investors, such positioning could strengthen FriendliAI’s value proposition to enterprises that require open-weight flexibility, multi-model portability, and production-grade reliability, potentially supporting customer acquisition and pricing power in the increasingly competitive AI infrastructure market.