Tech Times

Most AI Models Would Run Your Company Into the Ground, Princeton’s CEO-Bench Finds

Princeton’s CEO-Bench gave 14 AI models $1 million to run a simulated SaaS startup for 500 days. Most went bankrupt or lost ...
Tech Times

DeepSeek V4 Architecture: How Sparse Attention Cuts Inference Costs, What NIST Found

DeepSeek V4 architecture uses sparse attention to cut inference costs 73% at one-million-token contexts, but a NIST ...
西部e网

Hermes Agent 省钱指南：5 个便宜但能打的大模型

Nous Research 的 Hermes Agent 有个好处：你指向哪个模型，它就老老实实跑哪个。换句话说，账单多少是你自己配出来的，不是写死的。 所以选模型这件事，重点不在"哪个最聪明"，而在"哪个便宜模型够用"，以及"怎么配 Hermes，让它别白白烧 token"。 下面这五个模型都 ...
decrypt

DeepSeek Is Building Its Own Claude Code. Beijing Wants the Whole Stack

Add Decrypt as your preferred source to see more of our stories on Google. DeepSeek is recruiting a "Code Harness" team in Beijing to build a native agentic coding tool—a direct rival to Anthropic's ...
blockchain

DeepSeek V4 powers Claude Code integration, 97% cheaper

In the rapidly evolving landscape of artificial intelligence, recent advancements in model compatibility and cost optimization are reshaping how businesses access powerful AI tools. According to ...
GitHub

DeepSeek Code Worker MCP

Coding-worker MCP for Codex Desktop. Codex plans, delegates, and reviews; DeepSeek V4 does the expensive code reading, editing, and checking through Claude Code. Goal: save Codex main-thread tokens, ...
GitHub

yunshu0909/deepseek-claude-setup

让 Claude Code、OpenAI Codex CLI 和 Hermes Agent 透明使用 DeepSeek、智谱 BigModel、Kimi 模型，且思考模式真的生效。 当前版本（v1.6.5 ...
腾讯网

我让 Claude Code 当裁判，横评 DeepSeek V4 和 GLM-5.1

DeepSeek V4 发布时，在技术报告里写的非常真诚：在推理能力上，「落后前沿闭源模型大约 3 到 6 个月」。 而最近两天，我的各个编程交流群里就开始了对 V4 和各家模型进行大量比较、讨论。其中讨论到的国产模型，最多的一个我看下来就是智谱的 GLM 了： 两个 ...
VentureBeat

DeepSeek-V4 arrives with near state-of-the-art intelligence at 1/6th the cost of Opus 4.7 ...

On standard, cache-miss pricing, DeepSeek-V4-Pro comes in at roughly one-seventh the cost of GPT-5.5 and about one-sixth (1/6th) the cost of Claude Opus 4.7. With cached input, the gap widens: ...
CNBC

China's DeepSeek releases preview of long-awaited V4 model as AI race intensifies

DeepSeek launched a preview of its V4 model, expanding its open-source AI lineup. Its earlier R1 model disrupted markets with strong performance at a lower cost. Rising Chinese competition intensified ...
blockchain

DeepSeek V4 Integrates with Claude Code and OpenClaw: Latest Analysis on Agentic Coding ...

According to DeepSeek on X (Twitter), DeepSeek V4 is now natively integrated with leading AI agents including Claude Code, OpenClaw, and OpenCode, and is already powering in-house agentic coding ...
The New York Times

DeepSeek’s Sequel Set to Extend China’s Reach in Open-Source A.I.

Chinese companies have embraced making their most advanced artificial intelligence models available to all. The Chinese start-up DeepSeek shook the industry in January 2025 with its claim that it had ...