AITOPAI热报
ModelsSelected

Deepseek V4 Pro 0813实测：网站开发、60种设计风格、3D游戏

向阳乔木@vista8
Post
View the post
Why we picked it

vista8用Deepseek V4 Pro 0813跑了三个活：网站部署、60种设计风格、3D游戏，视频挺直观。

Deepseek V4 Pro 0813在网友vista8的测试中完成了三个任务。第一个任务调用3个Skill完成网站开发、子域名解析和上线。

Image · 向阳乔木
Body

第二个任务模仿60种设计风格生成Bento卡片图。第三个任务用Three.

js制作3D打砖块游戏。

Original · 向阳乔木

用Deepseek V4 Pro 0813跑了三个简单任务。 1. 第一个是开发部署一个网站，需要调用3个Skill，完成网站开发，子域名解析和上线。 2. 第二个是模仿60种设计风格，生成Bento卡片图展示。 3. 第三个是Threejs 生成一个3D打砖块游戏。 大家可以看看效果，我感觉还行。

Industry Coordinates & Source Verification

This report is categorized under AITOP's "Models" desk, focusing on China's AI industry and domestic foundation model ecosystem. Within our cross-regional analytical framework, it directly maps to US-China foundation model benchmarks and commercialization comparison. Key entities and technical topics under continuous tracking include Deepseek V4 Pro 0813, Three.js, 网站开发, 3D游戏.

Source attribution: curated from 向阳乔木 (@vista8).

#Deepseek V4 Pro 0813#Three.js#网站开发#3D游戏#设计风格

Keep reading

Recent stories with the same tag or section
ModelsBuilders81°13:58
Chinese contentSingularity-Bench 发布：测试智能体能否端到端构建 AI 系统

这个基准让智能体从零写训练、推理、编排系统，再和人类基线硬碰硬。结果挺扎眼：Opus 5.5 配 Claude Code 六项全过人类线，其他智能体平均只有 0.58。做 agent 评测或基础设施的工程师都该看看。

#Claude Code#Singularity-Bench#benchmark#智能体
arXiv cs.AI@Yang Zhou et al. (19)Source
ModelsBuildersMulti-source13:19
Chinese contentSaidzoda-Kazakh-EmbeddingGemma-300M：单轮微调将哈萨克语检索 recall@1 提升至 78.0%

339M 参数的小模型，用一个 epoch 的合成问答对微调，哈萨克语检索 recall@1 从 60.8% 拉到 78.0%，还压过 OpenAI text-embedding-3-large。做低资源语言检索或想复刻这套合成数据微调流程的工程师都该看看细节。

#EmbeddingGemma#MTEB#Belebele#嵌入模型
Event
arXiv: OpenAI@Tohir Saidzoda et al. (6)Source
ModelsBuildersSelected12:47
Decision Models for agentic security open sourced, free on OpenRouter

Open weights aimed at a niche most security teams have not tooling for yet: auditing agent sessions at full scale rather than sampling. Security and platform engineers operating multi-agent deployments can self-host under Apache or use the free OpenRouter window until 19 Nov to evaluate it against manual rule review. The Perplexity figure of 1 billion tokens and 14 rule fixes in a day gives a concrete benchmark for what such audit loops can yield.

#Decision Models#OpenRouter#Perplexity#agents
Aravind Srinivas@AravSrinivasSource