在国产大模型价格战里，DeepSeek 是最难绕开的名字。V3.2 标准输入 2 元、缓存命中 0.2 元、输出 3 元每百万 Token——这个定价在 Hacker News 和 Reddit ...

大家好，这里是程序员晚枫，2026继续all in AI。 我用AI编程快2年了，在AI编程课的交流群里，被问得最多的一个问题不是"怎么学AI编程"，而是： "晚枫，AI编程工具太多了，到底哪个便宜？哪个划算？Token去哪买最便宜？" ...

If you want to set and use Deepseek-R1 in Visual Studio Code, follow the steps below. Install Visual Studio Code Download Ollama Install the CodeGPT Extension Install DeepSeek models Use DeepSeek in ...

最近的 AI 编程圈，实在是太卷了。 感觉我就是睡了一觉，起来天就变了： 前脚Anthropic刚把Claude Code 升级到 2.0、发布Claude sonnet 4.5，后脚 DeepSeek 就密集更新到了 v3.2-exp。 还没等我缓过神，昨晚刷 X，智谱的 GLM-4.6 已经悄悄干到了开源第一，把 DeepSeek 给超了。

Last month, DeepSeek turned the AI world on its head with the release of a new, competitive simulated reasoning model that was free to download and use under an MIT license. Now, the company is ...

昨天，DeepSeek 发布了一个新模型 DeepSeek-OCR。 这是一个专门为 OCR（文字识别）微调的 6.6GB 模型，主要贡献在于首次量化 “视觉 - 文本 token 压缩比”，验证 10× 近无损压缩、20× 仍保有 60% 精度的可行性；提出 DeepEncoder，解决现有编码器 “高分辨率 - 低内存 - 少 ...

Chinese AI sensation DeepSeek plans to release key codes and data to the public starting next week, an unusual step to share more of its core technology than rivals such as OpenAI have done. The ...

WTF?! Chinese ChatGPT rival DeepSeek has been positively received since it shook the markets on its arrival in January. But the quality of code generated by the AI appears to depend on how China views ...