DeepSeek V4 Pro 0813 Max vs Qwen3.8 27B Xhigh compared

DeepSeek V4 Pro 0813 (Max) from DeepSeek and Qwen3.8 27B (Xhigh) from Alibaba are both tracked in Franklin AI's benchmark catalogue. Once text starts flowing, DeepSeek V4 Pro 0813 Max is the quicker writer by a wide margin, at 1.7 times the output speed of Qwen3.8 27B Xhigh. This page is written automatically from catalogue numbers only: benchmark scores, list prices and measured speed.

How fast they respond

DeepSeek V4 Pro 0813 Max generates about 67 tokens per second and Qwen3.8 27B Xhigh about 38 tokens per second. That makes DeepSeek V4 Pro 0813 Max roughly 1.7 times as fast at writing. Time to first token is 1.7 seconds for DeepSeek V4 Pro 0813 Max and 4.1 seconds for Qwen3.8 27B Xhigh. DeepSeek V4 Pro 0813 Max gets to its first token 2.4 times as fast.

Put together, a 2,000-token answer would take about 32 seconds from DeepSeek V4 Pro 0813 Max and 56 seconds from Qwen3.8 27B Xhigh, counting the first-token wait.

Reasoning and overall scores

DeepSeek V4 Pro 0813 Max comes out ahead on the intelligence index, 23.1 to 20.1. That is a clear margin for an index that averages many evaluations. Put another way, Qwen3.8 27B Xhigh would need 15% more on this index to draw level.

Among 109 active scored models, DeepSeek V4 Pro 0813 Max sits 72nd and Qwen3.8 27B Xhigh 78th.

Specialist coding and math indexes are missing for both, so treat the overall score as the only capability signal here.

Generation gap

DeepSeek V4 Pro 0813 Max came first, on August 13, 2026; Qwen3.8 27B Xhigh followed on August 14, 2026. They are effectively contemporaries.

Price per token

Per million tokens, DeepSeek V4 Pro 0813 Max charges $1.32 for input and $3.96 for output, while Qwen3.8 27B Xhigh charges $0.50 and $3.00. For output tokens, DeepSeek V4 Pro 0813 Max is 1.3 times the price of Qwen3.8 27B Xhigh.

For a sample workload of 1 million input tokens and 200,000 output tokens, DeepSeek V4 Pro 0813 Max would cost $2.11 and Qwen3.8 27B Xhigh $1.10, a difference of $1.01. A bigger example, a month of coding-assistant traffic at 30 million input and 3 million output tokens, costs $51.48 on DeepSeek V4 Pro 0813 Max and $24.00 on Qwen3.8 27B Xhigh: a gap of $27.48. On a 3:1 blend of input to output tokens, that works out to $1.98 versus $1.125 per million.

Cost per point of intelligence

Dividing the blended price by the intelligence index gives a rough cost per point: $0.086 for DeepSeek V4 Pro 0813 Max and $0.056 for Qwen3.8 27B Xhigh, per million tokens. So DeepSeek V4 Pro 0813 Max buys its higher score at a premium, while Qwen3.8 27B Xhigh is the value pick, paying 35% less per point.