MoonshotAI
Featured
Kimi-K3
Kimi K3 is an open-weight, 2.8T-parameter native multimodal agentic model with a 1M-token context window, designed for frontier coding, knowledge work, and reasoning tasks.
deepseek-investor-meeting
This repository hosts the complete transcript of a 3h44m investor meeting by DeepSeek founder Liang Wenfeng, covering AGI vision, open-source philosophy, compute/domestic chip strategy, and organization management, reorganized into 5 thematic Markdown files plus key-point summaries.
Stars
9
7-day growth
No data
Forks
3
Open issues
0
License
No data
Last updated
2026-07-27
AI repository intelligence
FR-AI / ANALYSISWhy it is worth attention
It provides an original, first-hand record of DeepSeek’s strategic thinking from its founder, including rare details on compute gaps, CUDA disruption via TileLang, continual learning as the next bottleneck, and their profit-driven API pricing model—all in a searchable, timestamped format.
Who it is for
- AI investors and venture capitalists
- Deep learning researchers and engineers
- Chinese tech startup founders and product managers
- Policy analysts tracking Sino-US AI competition
Use cases
- Studying DeepSeek’s AGI roadmap and open-source rationale
- Understanding the current state of domestic chip alternatives (Huawei 950 vs NVIDIA)
- Extracting organizational management insights from a vision-driven AI lab
- Benchmarking API pricing strategy (6x profit, 10-month hardware payback)
Strengths
- Complete verbatim transcript with precise timestamps for easy citation
- Dual format: full transcript (5 files) and condensed bullet-point summaries (in /纪要 folder)
- Merges two separate GitHub archives into a single, well-organized repo
- Covers all major topics: technical roadmaps, compute economics, team culture, investor Q&A
Considerations
- Auto-transcribed from speech recognition, so proper nouns and numbers may contain errors (original audio required for verification)
- Speakers are not differentiated, making it hard to attribute questions vs answers
- Represents only a single event (May 2026); views may not reflect current DeepSeek positions
README quick start
梁文锋四小时投资人会议实录｜DeepSeek 投资者交流会完整文字稿
2026 年 5 月 20 日，DeepSeek 创始人梁文锋在一场持续约 3 小时 44 分钟的投资者交流会上，系统讲述了 DeepSeek 的 AGI 愿景、开源哲学、算力与国产芯片判断、组织管理思路，并回答了投资人关于持续学习、Coding Agent、中美 AI 差距、资本市场平衡等热门问题。本仓库即这场梁文锋四小时投资人会议实录的完整录音文字稿，按主题重新整理为 5 个 Markdown 文件，方便检索与引用。
一、实录简介
- 会议名称：DeepSeek 投资者交流会（业内常被称为"梁文锋投资人会议"）
- 录音日期：2026 年 5 月 20 日
- 整理日期：2026 年 7 月 16 日
- 音频时长：约 3 小时 44 分钟（故外界通称"四小时会议"或"梁文锋四小时会议"）
- 主讲人：梁文锋（DeepSeek 创始人）
- 内容形式：语音识别自动转写 + AI 整理，保留原始时间戳
[HH:MM:SS]，未区分说话人 - 本仓库定位：梁文锋四小时投资人会议实录的首发完整 Markdown 版本，按主题切分，支持全文检索
个别专名与数字可能存在识别误差，引用请以原录音为准。
二、目录结构
本仓库将完整的梁文锋四小时投资人会议实录拆分为 5 个主题文件，每个文件聚焦 1–2 个核心议题，时间戳在文件内按时间顺序排列：
| 文件 | 主题 | 覆盖时段（音频） |
|---|---|---|
DeepSeek梁文锋四小时投资者交流会-01-愿景与开源.md | 公司愿景、开源哲学、克制战略、合理利润与 API 定价逻辑 | 00:00:01 – 02:34:59（散段） |
DeepSeek梁文锋四小时投资者交流会-02-AGI技术路线.md | AI 发展阶梯（语言模型 → CoT → Agent → 持续学习 → 奇点 → 具身智能）、模型规划、多模态、Scaling | 00:51:02、02:08:18 |
DeepSeek梁文锋四小时投资者交流会-03-算力与国产芯片.md | 中美算力差距、CUDA 生态瓦解、TileLang、华为 950 与国产替代 | 01:26:41、01:56:36 |
DeepSeek梁文锋四小时投资者交流会-04-组织管理与商业化.md | 愿景驱动组织、共识决策、团队稳定性、竞争格局、商业化路径 | 01:02:45、01:39:46 |
DeepSeek梁文锋四小时投资者交流会-05-投资者Q&A.md | 投资者自由提问环节完整记录：持续学习、AI 硬件短板、组织架构、产品路线、资本市场平衡等 | 02:34:59 – 03:44:00 |
纪要版（要点概括）
纪要/ 文件夹存放的是按主题分类的发言要点整理版（非逐字实录），适合快速浏览核心观点：
| 文件 | 说明 | 来源 |
|---|---|---|
纪要/梁文锋四小时投资人会议实录-Long版.md | 118 条发言要点，按 01-10 主题分类，仅略作编辑 | 0xtresser 仓库 |
纪要/梁文锋四小时投资人会议实录-elsewhere版.md | 另一版本的要点整理 | 0xtresser 仓库 |
纪要/梁文锋投资者交流会-录音.pdf | 原始录音转写 PDF（1.2MB） | 0xtresser 仓库 |
完整版 vs 纪要版：根目录的 5 个 md 文件是逐字录音转写稿（带
[HH:MM:SS]时间戳，共约 108KB）；纪要/文件夹是二次概括的要点整理（约 42KB）。想看原话用完整版，想快速抓重点用纪要版。
仓库来源说明
本仓库为整合版，合并了 GitHub 上两个公开仓库的内容，整理为单一归档：
- 完整版主体（根目录 5 个 md）：来自
iamsophie/deepseek-liang-wenfeng-investor-meeting（逐字录音转写稿） - 纪要版（
纪要/文件夹）：来自 [0xtresser/Transcript-of-Liang-Wenfengs-DeepSeek-Founder-4-Hour-Investor-Meeting](https://github.com/0xtresser/Transcript-of-Liang-Wenfengs-DeepSeek-Founder-4-Hour-Investo
Description
DeepSeek梁文锋投资者交流会完整实录+纪要整合版（2026-05-20，约3h44min）
Related repositories
Similar projects matched by category, topics, and programming language.
xuchonglang
Featured
investing-for-beginners
A structured investing guide for Chinese beginners covering US stocks, options, and cryptocurrency, with focus on foundational concepts and risk awareness.
Krishnagangwal
Featured
CS-Fundamentals
A curated collection of Computer Science fundamentals (PDFs, notes, cheatsheets, interview question banks) for placement preparation, covering seven core subjects plus general resources.