# DeepSeek-V4-Pro 以低价逼近 GPT-5.6

- 来源：Rohan Paul (@rohanpaul_ai)
- 发布时间：2026-08-13 16:06
- AIHOT 分数：52
- AIHOT 链接：https://aihot.virxact.com/items/cmsr90wgl0ps0roz2m49tmhah
- 原文链接：https://x.com/rohanpaul_ai/status/2087813049537401235

## AI 摘要

DeepSeek-V4-Pro (Max) 在 Code Arena WebDev AutoEval 获 1607 分，仅落后 GPT-5.6 Sol xHigh 15 分，但混合 token 价格约为其 1/31。该开源权重模型定价 $0.435/$0.87 每百万 token，超越 Opus-4.8 与 GLM-5.2 等高价竞品，位列总榜约第 8。

## 正文

DeepSeek-V4-Pro (Max) is expected to shift the Pareto frontier

DeepSeek-V4-Pro (Max) reached 1607 in Code Arena WebDev AutoEval at unusually low API pricing.

So DeepSeek sits only 15 Arena points behind GPT-5.6 Sol xHigh while charging about 1/31st of its blended token price.

Kimi K3 Max leads it by 67 points at 1674, but its listed $3/$15 pricing produces a blended cost almost 16 times higher.

Code Arena: WebDev tests models on building interactive web applications,

### 引用推文

> Arena.ai：DeepSeek-V4-Pro (Max) by @deepseek_ai is expected to shift the Pareto curve for Code Arena: WebDev with this upcoming open weights model. It currently sits at ~...