Mercury 2.5: LLM model achieves 770 tokens per second

The Mercury 2.5 model achieves 770 tokens per second in inference speed.

Published 2h1 sourceNotable

≈ 17sSpeed

The fact

This performance is measured by Artificial Analysis using standardized benchmarks.

The model aims to optimize costs and latency for AI applications.

Click the link to read an article on the topic:

Explore this topic

Auto-synthesis from 1 media source · identified on September 24, 2026