Qwen3.7 Max ties GPT-5.4 in benchmark, trails by 6 points vs flagship
Alibaba's open-weight model matches top-tier rivals, smaller 27B variant lags behind.
Deep Dive
Qwen3.7 Max sits at 5th, on par with GPT 5.4 (xhigh) and a notch above Gemini 3.5 Flash. Meanwhile, Qwen3.6 27B trails exactly 6 points behind its max counterpart.
Key Points
- Qwen3.7 Max ranks 5th in Artificial Analysis benchmark, tying GPT-5.4 (xhigh) and beating Gemini 3.5 Flash.
- Smaller Qwen3.6 27B model scores exactly 6 points lower, indicating a clear performance-size tradeoff.
- DSV4 Flash (DeepSeek V4 Flash) also appears as a competitive open-weight entry on the leaderboard.
Why It Matters
Open-weight models now rival top proprietary AI, cutting enterprise costs for high-end reasoning tasks.