Create

Sign in to ReadmeX

or

How we handle your data: Privacy Policy

ReadmeX

Your AI reading companion

What would you like to catch up on?

Pick a briefing, or ask your own question.

Your community and people briefings stay personal to you.

DeepSeek

AI overview

Sign in and the AI will write an overview from our coverage.

Headlines · 471

  1. DeepSeek Details a Super-agent Architecture for Continuous, Concurrent Work

    A published analysis explains DeepSeek’s Super-agent design, which separates routing from execution into Router and Worker layers to support continuous cross-channel interaction and concurrent tasks. It also describes fast and slow paths, same-thread direct routing, Inline and Spill context handling, asynchronous execution, and an external Watchdog. DeepSeek’s ablation results reportedly cut Router judgments by about 30% and repeated KV Cache token reads from 31.77 million to 23.23 million, though independent validation was not provided.

    掘金 · 人工智能 · 🔥 4
  2. Tutorial Outlines a Gateway Pattern for Switching AI Model Providers

    An article in the “AI Application Production Handbook” presents a three-layer architecture in which applications use logical model names, the gateway maps those names to provider-specific models, and providers such as DeepSeek and OpenAI sit behind the gateway. Its vendor_switch.py examples check configuration consistency, scan application code for hard-coded provider names, and simulate a DeepSeek-to-GPT-4o migration without changing application code. The article emphasizes that a production switch also requires evaluation against a golden set, a quality threshold, staged rollout, A/B testing, and rollback to a backed-up configuration if evaluation fails.

    掘金 · 人工智能 · 🔥 3
  3. Build a Poetry-Aware Homework Assistant with LangChain4j RAG

    A Juejin tutorial shows how to add an eight-poem private corpus to the “字问先生” homework assistant using LangChain4j RAG. The system embeds each poem, retrieves semantically relevant entries, injects them into the prompt with title, author, dynasty, and lesson metadata, and asks the model to abstain when no matching poem exists. The reported accuracy and anti-hallucination behavior have not been independently benchmarked.

    掘金 · 人工智能 · 🔥 3
  4. Building a One-Minute AI Short with Seedance 2.5 Flash

    A first-person developer workflow shows how a father used DeepSeek, Doubao image generation, DeepSeek Harness, and Seedance 2.5 Flash to turn his child’s drawings into a one-minute superhero short. The author split the story into two clips, used reference images and a light-based transition to connect them, then generated and combined the videos through Python. The author reports occasional prop-consistency errors and unclear native speech, giving the result a subjective score of 8 out of 10.

    掘金 · 人工智能 · 🔥 3
  5. Set Up a LiteLLM AI Gateway for Model Routing and Key Control

    This tutorial shows how to put LiteLLM Proxy between an application and model vendors such as DeepSeek and OpenAI. It covers logical model names, virtual keys, fallback, rate limits, PostgreSQL-backed spend tracking, and container-based validation of health, model listings, key creation, deletion, and spend logs. The author also compares LiteLLM with hosted Portkey and observability-focused Helicone.

    掘金 · 人工智能 · 🔥 2
  6. DeepSeek reportedly limits free chat usage and long conversations

    Users reported that DeepSeek's free chat service began showing usage-limit notices, with some long conversations becoming unavailable and other sessions hitting limits in under two hours. According to the republished APPSO report, conversations inactive for more than seven days can no longer continue, single requests are limited to roughly 40,000 characters, and daily usage caps trigger a “server busy” notice until the next day. The limits were not confirmed by an official DeepSeek announcement in the supplied report.

    36氪 人工智能 · 🔥 3
  7. Volcano Engine Launches AI Data Lake LAS

    Volcano Engine has launched LAS, a fully managed enterprise data foundation cloud service designed for AI and multimodal workloads. The product integrates multimodal operators, lake storage, dataset governance, elastic compute and Agent interfaces, including MCP Server, CLI, API, SDK, Skills and WebUI, while supporting data flow to Volcengine Ark and its machine learning platform. The report says LAS includes more than 150 online and offline multimodal operators covering text, images, audio, video, documents and embodied-sensing data.

    掘金 · 人工智能 · 🔥 1
  8. AI App Shutdown Wave Coincides With Rising Consumer Demand

    钛媒体 and 36氪人工智能 published the same analysis of AI product closures, including Sora, 冒泡鸭, XEVA, 字节扣子企业版, 腾讯元器, 百度梯子AI and 米哈游 AnuNeko. The reports say Dang.ai had marked 1,895 discontinued tools and services by November 20, 2026, while Sensor Tower counted 3.8 billion generative-AI mobile-app downloads in 2025 and more than $5 billion in in-app purchase revenue. The analysis attributes closures to pressure from rising inference costs, feature convergence, limited differentiation and changing product portfolios, rather than to an overall collapse in demand.

    36氪 人工智能 · 🔥 6
  9. Lenovo Baiying Launches Token Plan and Token Key

    Lenovo Baiying introduced the Token Plan platform and Token Key hardware solution to connect model access, subscription entitlements, and endpoint applications. Token Plan provides model packages, usage queries, and centralized management of members, credentials, and quotas, while the Alibaba Cloud-based architecture adds traffic routing, fallback, metering, and protocol adaptation. Lenovo Baiying describes Token Key as China's first integrated software-and-hardware Token solution, but that claim was not independently verified.

    掘金 · 人工智能 · 🔥 1
  10. FutureSearch Ranks AI Fortunes, Led by Liang Wenfeng at $44B

    FutureSearch estimated the wealth of 60 AI-linked millionaires, with the top ten ranging from $12 billion to $44 billion; TechNews ranked DeepSeek founder Liang Wenfeng first at $44 billion. The report says many fortunes are concentrated paper wealth tied to company valuations, share prices, litigation or IPO outcomes, while investors are reportedly discussing a funding round that would value DeepSeek at $74 billion.

    科技新报 · 🔥 3
  11. Tencent Cloud waives DeepSeek Search and Fetch fees

    Tencent Cloud announced that, starting at 12:00 p.m. Beijing time on November 18, 2026, server-side Search and Fetch triggered by calls to DeepSeek original-factory directly supplied models will no longer be charged. Calls will be billed only for model Token usage, and Tencent Cloud will refund eligible server-side internet-search fees incurred after that time by November 30, 2026, according to 36Kr.

    36氪快讯 · 🔥 3
  12. DeepSeek Reportedly Adds Free-User Limits and 7-Day Conversation Expiry

    DeepSeek free users reportedly found that conversations inactive for more than seven days can no longer be resumed, while single inputs are capped at about 40,000 Chinese characters and daily usage may trigger quota or server-busy prompts. The report attributes the constraints to insufficient computing capacity, but the limits were not independently verified and the capacity claim was not supported by published data.

    虎嗅 AI · 🔥 5
  13. DeepSeek Harness v0.1 Open-Source Developer Preview Released

    DeepSeek reportedly released the DeepSeek Harness v0.1 developer preview, abbreviated dsh, on August 13, 2026, alongside open-source code. The Juejin AI article describes Harness as an agent runtime layer that connects models with tools, sessions, permissions, storage, and interfaces, rather than as a new DeepSeek large model. Its architecture includes a replaceable model provider and plugin-based tools, but the report does not independently verify compatibility with other model providers.

    掘金 · 人工智能 · 🔥 0
  14. DeepSeek Tightens Free-User Limits on Web and App

    TMT Media reports that DeepSeek has tightened usage limits for free users of its web service and app. Long conversations inactive for more than seven days can no longer be continued, the text input in a new conversation is limited to roughly 40,000 Chinese characters, and about two hours of continuous use triggers a “server busy, usage limit reached” notice. The report also discusses DeepSeek’s paid APIs and presents inference costs as a factor that makes unlimited free use harder to sustain.

    钛媒体 · 🔥 0
  15. Tsinghua Yao Class alumni form a prominent global AI network

    A 36Kr article republished from Phoenix Weekly uses Tencent chief AI scientist Yao Shunyu’s November 18 lecture at Tsinghua’s Yao Class to examine the school’s role in AI talent. It identifies alumni in AI research and leadership roles at Tencent, DeepSeek, OpenAI, Meta and xAI, as well as in academia, infrastructure and applied technology. The report says claims that Yao Class alumni earn exceptional salaries, including claims about Yao Shunyu’s Tencent compensation and DeepSeek internships, have been debunked.

    36氪 人工智能 · 🔥 0
  16. Tencent releases Hy4 multimodal MoE model with 1M-token context

    The Tencent Hy Team introduced Hy4, a new-generation Mixture-of-Experts flagship model for text, image, and video understanding, according to its official model card. Its 770B-parameter language backbone activates 49B parameters per token and supports a 1M-token context, while a 3.3B DFly2 drafter is included for speculative decoding. Tencent also provides FP8 and GGUF artifacts, vLLM and SGLang deployment recipes, a finetuning pipeline, and documentation under the Tencent Hy License.

    腾讯混元 · Hugging Face · 🔥 0
  17. Developer opens state-driven companion runtime as npm package

    The project author open-sourced @yanisjiang/companion-core, a TypeScript runtime for adding state-driven companion behavior to existing applications. According to the author, traits and current state determine an intent, silence can bypass the LLM, and developers can connect the package to bots, game NPCs, desktop shells, or custom backends by supplying an LLM and storage. The repository’s Web/PWA implementation is presented as an optional example host rather than part of the core package.

    掘金 · 人工智能 · 🔥 0
  18. Ten Silent Runtime Failures in a DeepSeek Coding Agent

    A developer describes six months of operating a custom terminal-based coding agent backed by DeepSeek, with 359 commits reported, including 129 bug-fix commits. The article details ten runtime failures that appeared normal, including missing completion signals, discarded input, false success, context compression losing task anchors, and cache-prefix invalidation. It recommends explicit failures, machine-readable command results, queued input, stable context anchors, and deterministic cache prefixes, but the usage and performance figures are self-reported.

    掘金 · 人工智能 · 🔥 0
  19. Client-Side Concurrency Cuts DeepSeek Agent Tool Latency

    A developer reported that changing 12 tool calls in an internal code-inspection agent from serial to concurrent execution reduced elapsed time from 94 seconds to about 23 seconds. Their trace attributed 71 seconds of the original run to tool and API waits and only eight seconds to model inference. The Juejin AI report details client-side scheduling, concurrency limits, exception handling, result validation, streaming aggregation, and a two-stage context strategy that reduced input from 21,000 to 5,100 tokens.

    掘金 · 人工智能 · 🔥 0
  20. Tony Zhang open-sources Linger personal AI agent

    Simplex AI founder Tony Zhang has open-sourced Linger, a personal agent designed to run on a user's computer, reuse existing browser and application sessions, and connect third-party or local models. Its Another Me component can monitor email, calendars, chats, and meetings, maintain a local knowledge base, issue proactive reminders, and learn notification preferences from feedback; the project also includes the PolyStrata local inference engine. Vendor-reported claims include 95% accuracy on LoCoMo and local inference of Qwen3.8-Flash-Next on consumer hardware, while GitHub stars reportedly exceeded 2,000 shortly after launch.

    量子位(原生 RSS) · 🔥 8
  21. Lightspeed Partner Reflects on Anthropic Bet and AI Competition

    Lightspeed Venture Partners partner Sebastian Duesterhoeft said backing Anthropic became one of the firm’s boldest bets as generative AI rapidly expanded. In a Bloomberg interview, he discussed lessons from the DeepSeek moment, Anthropic’s focus on developing AI safely, and why competition among multiple leading AI models could ultimately strengthen the industry.

    Bloomberg Technology · 🔥 0
  22. Baseten Adds Deferred Tool Loading for Open Models

    Baseten says its Model APIs now support deferred tool loading for DeepSeek, Kimi, and GLM in the Chat Completions and Messages API formats. In its benchmarks using Claude Code, coding tasks were on average 62.5% cheaper and 26.2% faster, while agentic tasks were 51.1% cheaper and 14.9% faster. The tests used 55 tasks across eight benchmarks. Baseten reported that DeepSWE and MCP-Atlas tasks were slightly slower, while eager loading was more efficient on Toolathlon-GYM, where tasks averaged 68 tool calls.

    Baseten Blog · 🔥 0
  23. ThinRead 0.2.2 retrospective details a local-first AI reading workflow

    In a first-party retrospective, the author describes ThinRead 0.2.2, an MIT-licensed macOS reading-and-writing application built with Electron, React and TypeScript. The tool stores books and notes locally, verifies generated quotations against source text, validates streamed AI output before writing, and uses local BGE-small-zh embeddings for cross-book connections, while supporting multiple cloud and local model providers. The author also reports an import pipeline covering eight file formats, webpages, OPDS, WeChat Reading, Readwise and Kindle highlights, plus macOS OCR for scanned PDFs.

    掘金 · 人工智能 · 🔥 0
  24. AI Agent Short-Term Memory: Storage, Truncation, and Compression

    A report on AI agent short-term memory presents a three-stage engineering approach: storage, truncation, and compression. It shows how to persist conversation history in JSON, pass only a recent-message window to an LLM, and use an LLM-generated summary to replace older messages. The compression workflow is designed to retain user identity, important decisions, technical details, and unfinished tasks, while keeping the injected message history within context limits. The examples use OpenClaw’s session-management pattern.

    掘金 · 人工智能 · 🔥 0
  25. Build a local RAG assistant with Xiaoji v0.5

    This fifth installment of the 手搓 Agent series explains retrieval-augmented generation through the example project “小记,” which the author upgrades to v0.5. It demonstrates embedding-based search with bge-small-zh-v1.5, chunking, hybrid vector and keyword retrieval, reranking concepts, and exposing search as an agent tool. The author reports that one example query triggered four searches and produced an answer with sources, while also noting that results vary by model and run.

    掘金 · 人工智能 · 🔥 0
  26. DeepSeek Adds Hangzhou Chengyan as Shareholder

    Registry records reviewed by 36Kr, Zhidx and IT Home show that Hangzhou Chengyan became a shareholder of DeepSeek operator Hangzhou DeepSeek, whose registered capital rose from about RMB 16.45 million to RMB 19.92 million. Hangzhou Chengyan added 31 partners, including companies linked to Tencent, CATL, China Mobile, China Telecom, NetEase, Midea and Geely, while founder Liang Wenfeng left its direct partner list. Hangzhou Chengce, which Liang controls, remains its executive partner.

    智东西 · 🔥 3
  27. DeepSeek Capital Increase Reported at About 19.92 Million

    A report states that DeepSeek's capital increased to approximately 19.92 million. The available report does not specify the currency, investors, or any other details about the increase.

    DoNews · 🔥 0
  28. Who Should Bear Responsibility for “AI-Flavored” Writing?

    A commentary from the WeChat account 405游局 uses media commentator Hu Xijin’s leftover AI interaction phrase to examine public distrust of AI-assisted writing. Hu said the post was dictated and organized with AI, but that the views were his own. The commentary notes that repetitive phrasing, excessive symmetry, and familiar transitions may prompt suspicion of AI involvement, yet usually do not establish whether a text was generated, polished, or otherwise assisted by AI. It distinguishes model limitations from an author’s responsibility for selecting, revising, and publishing the final work.

    虎嗅 AI · 🔥 0
  29. Kimi Developer's Yunqiao Appearance Fuels AI Persona Debate

    Kimi engineer and developer-relations lead Tang Feihu appeared at the 2026 Yunqiao Conference wearing a DeepSeek whale costume. He said Kimi's own costume was not ready and that wearing women's clothing and participating in cosplay had been part of his life for more than a decade. In the interview, he described harnesses as evolving transitional tools, argued that open-source ecosystems benefit from diverse technical paths, and said models should challenge more fundamental benchmarks. A TMT commentary used the appearance to argue that Kimi has not developed a recognizable community-created AI persona comparable to DeepSeek's whale character; this was commentary, not an official Kimi position.

    36氪 人工智能 · 🔥 0
  30. DeepSeek Reportedly Limits Conversations for Ordinary Users

    One report states that DeepSeek has imposed conversation quotas for ordinary users. A commentary republished by Huxiu characterizes the change as rate limiting and attributes it to insufficient computing capacity, but notes that the view represents the author’s opinion rather than Huxiu’s position.

    DoNews · 🔥 0

Next →

Experience and discussion from the community

Share my DeepSeek experienceAsk about DeepSeek

go-stock: A desktop stock analysis tool that combines market data with large language models.

What problem it solves Stock analysis often requires users to gather market quotes, news, technical indicators, financial information, and AI-assisted interpretation across multiple sources. The project addresses this by bringing supported-market data, analysis tools, AI agents, alerts, and review workflows into a locally focused desktop application. Key features Multi-market market data — The app…

github.comhttps://github.com/ArvinLovegood/go-stock

opencodex: A lightweight local gateway for using chosen models with AI coding tools.

What problem it solves AI coding tools often expect particular provider protocols, making it difficult to use the models and providers a user prefers. A local gateway can translate between supported client APIs and provider APIs while preserving capabilities such as streaming, tool calls, reasoning, and images. Key features Provider protocol translation — The gateway translates Responses, Chat Com…

github.comhttps://github.com/lidge-jun/opencodex

system_prompts_leaks: A regularly updated archive of leaked system prompts from ChatGPT, Claude, Gemini, and other AI products.

What problem it solves The hidden system instructions that shape AI product behavior are difficult to inspect and change across models, products, and versions. This repository makes collected prompts available in a structured, product-oriented archive for study and comparison. Key features Broad product coverage — The archive covers prompts from major AI providers and products, including Anthropic…

github.comhttps://github.com/asgeirtj/system_prompts_leaks

dsh-image-gen: An AI image studio for DeepSeek Harness that generates, edits, animates, and compares images through chat.

What problem it solves Image creation often requires switching between separate tools, manually translating visual ideas into parameters, and losing the context of an ongoing conversation. This project brings generation, editing, reference workflows, model comparison, and asset management into a DeepSeek Harness conversation and canvas. Key features Chat-Based Image Creation and Editing — Describe…

github.comhttps://github.com/shanliuling/dsh-image-gen

Parallel Agent Teams: Small-Team Collaboration, Multi-Team Exploration

Optimizing a GPU kernel or completing a set of formal proofs often requires repeated trial and error: try a different implementation, run an evaluation, and keep iterating in whichever direction proves effective. When giving a model more test-time compute, the most direct approach is to let more agents try simultaneously; if the task only needs one best result, the value of the other attempts lies…

AgentPAT测试时计算多智能体Benchmark

arxiv.orghttps://arxiv.org/abs/2611.09798

TradingAgents: A multi-agent trading framework that mirrors the dynamics of real-world trading firms.

What problem it solves Complex trading decisions require evaluation across financial fundamentals, market sentiment, news, technical indicators, portfolio holdings, and risk. TradingAgents addresses this by decomposing trading tasks into specialized LLM-powered roles that discuss findings and assess proposed transactions. Key features Parallel analyst research — Fundamental, sentiment, news, and t…

github.comhttps://github.com/TauricResearch/TradingAgents

ds-harness-remote: Access your local Harness instance remotely from the web, a PC, or a mobile device.

What problem it solves DeepSeek Harness workspaces, tools, and project setup normally remain tied to the computer running the Harness instance. The project addresses the need to continue using that environment from another computer, phone, or browser without requiring a public listening port or public IP on the Host. Key features Remote workspace access — Open local workspaces from a web browser, …

github.comhttps://github.com/liguobao/ds-harness-remote

Evolver: Let Exploration Leave Behind Strategies

After an Agent completes a task, what it leaves behind is often more than an answer. There is also an expensive exploration path: which files it read, which tools it tried, where it failed, and what checks it used to confirm the result. The problem is that the next execution may not be able to use those experiences. The model parameters have not changed, and a new session will not naturally inheri…

EvolverGEPEvoMapEvolutionSkill

0x01

0x01 At the end of 2022, ChatGPT exploded onto the scene, triggering the first wave of excitement in the tech industry. But traditional state-owned enterprises did not act immediately. The real action began in 2023. Domestic large models were open-sourced one after another, and integrated hardware-and-software delivery formats such as “large-model all-in-one machines” became the default option. Mo…

RSI: Evolution Goes Beyond Models

AI has already entered code writing, algorithm optimization, and model development. But after a task is completed, what remains may be reusable capability—or merely an answer and scattered logs. The difference becomes apparent only when a similar problem comes up again: can the system reuse what it learned, or must it pay the cost of exploration all over again? Recursive Self-Improvement (RSI) co…

RSISwarmEvoMapLoopEvolution

evomap.aihttps://evomap.ai

Implementing Local Development, Git Commits, and Automated Server Deployment with DeepSeek Harness

My current development workflow for my own projects is to use agent tools like deepseek harness locally to write code and make git commits, then install deepseek harness on the server and connect to it through a tunnel, allowing the server-side deepseek harness to handle deployment. The server is Ubuntu, and the local Mac establishes the tunnel connection. Deployment only requires sending a deploy…

DeepSeek HarnessGitMac

Writing code with AI isn’t hard; accepting it is

First, the result: over the course of four months, on and off, I built a WeChat Mini Program called “随了么小管家” for keeping track of gift-giving and reciprocity. It currently has about 40 pages, with roughly 29,000 lines of frontend code and 14,000 lines of backend code, 85 commits, and a version number that has reached 1.1.9. The tech stack is pretty ordinary: uni-app + Vue 3 + Pinia on the fronten…

MySQLExcel通义千问QwenCodex