Against Qwen3.7 Max, Qwen reports gains across long-horizon and visual agent benchmarks: PaperBench rises from 64.8 to 93.0, OSWorld-Verified from 73.3 to 86.1, and Vision2Web from 42.1 to 69.0.

Qwen3.8 Max by @Alibaba_Qwen is live on OpenRouter. The new flagship has 2.4T parameters (95B active) and is built for long-horizon coding, research, and multimodal agent work. Open weights are due next week, a first for a Qwen Max-class…

want to give a huge shoutout to my team for an incredible july. our velocity continues to increase and i'm excited about our expansion into multimodality.

Personal evals will be quite important going forward. Love to see the OpenRouter team doing this. Hats off!

We spend a ton of time with customers struggling to write evals and figure out the best model for each task. This CLI wraps that experience into a single tool, allowing you to leverage all 500+ models on OpenRouter, their benchmarks, and real-world usage data. It's brand…