modelgrep

Qwen: Qwen3.7 Max vs Tencent: Hy3

qwen logoQwen: Qwen3.7 Maxqwen/qwen3.7-maxVStencent logoTencent: Hy3tencent/hy3

Qwen: Qwen3.7 Max wins on more metrics (5 of 7), but the right pick depends on what you optimize for — see the breakdown below.

MetricQwen: Qwen3.7 MaxTencent: Hy3
Intelligence Index46.0✓41.2
Coding Index66.0✓58.8
GPQA Diamond92%✓90%
Design Arena Elo1324✓1198
Speed (tokens/sec)——
Latency——
Input price /M$1.25$0.140✓
Output price /M$3.75$0.580✓
Context window1M✓262K
CapabilitiesReasoningToolsJSONReasoningToolsJSON
Qwen: Qwen3.7 Max details →Tencent: Hy3 details →