AI BriefingQwen 3.6 Comparison
Key point
In a 2 x RTX Pro 6000 environment, Qwen3.6-27B delivered the best performance, while Qwen3.6-35B-A3B showed higher tps.
Details
While running an on-premise LLM in a 2 x RTX Pro 6000, FP8, MTP environment and sequentially applying Qwen3.5-122B, Qwen3.6-35B-A3B, and Qwen3.6-27B, the perceived performance order was Qwen3.5-122B < Qwen3.6-35B-A3B < Qwen3.6-27B.
Speed and context figures were presented as follows.
- Qwen3.6-35B-A3B: 512k x 11, 280 tps
- Qwen3.6-27B: 320k x 6, 110 tps
For handling work requests, Qwen3.6-27B was sufficiently stable, while the reported throughput was higher for Qwen3.6-35B-A3B.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.
Related
Cerebras Serves Qwen 3.8 27B Model at 1,500 Tokens per Second
Apple Releases LensVLM-9B for Processing Compressed Text Images
Qwen Releases 'RecreationWorld' for Training AI Agents to Rebuild AppsQwen Intelligence Launches Three Mobile AI Agents and Open Benchmark Suite