# Kapyn

> Kapyn is the calm map of the AI worth using, agents, models, tools, MCP servers and skills, curated and kept current by a daily signal. It also distills each day's AI/tech news into 30-second dispatches, and auto-generates source-grounded explainers for the concepts behind that news. No paywall.

## Key pages
- [Kapyn](https://kapyn.app): the curated map, plus the swipeable daily news feed
- [Radar](https://kapyn.app/radar): curated AI agents, models, tools, MCP servers & skills
- [MCP servers](https://kapyn.app/mcp): directory of Model Context Protocol servers, by category
- [AI tools](https://kapyn.app/tools): the essential AI tools worth knowing, by category
- [AI skills](https://kapyn.app/skills): custom GPTs, Claude Skills & Gemini Gems, by use case
- [Compare AI models](https://kapyn.app/compare): Claude, GPT, Gemini, Llama and more side by side , context, cost, modalities, best-for
- [Explore , the AI glossary](https://kapyn.app/explore): index of every concept explainer
- [Blog](https://kapyn.app/blog): guides on the AI and tools worth using

## AI concepts & techniques
- [AI Agents](https://kapyn.app/learn/agents): LLM systems that plan, call tools, and act over multiple steps toward a goal.
- [Model Context Protocol](https://kapyn.app/learn/mcp): An open standard for connecting AI assistants to external tools and data sources.
- [Retrieval-Augmented Generation](https://kapyn.app/learn/rag): A technique that grounds an LLM's answers in documents retrieved at query time, reducing hallucination.
- [Mixture of Experts](https://kapyn.app/learn/mixture-of-experts): A model design that routes each input to a subset of specialised sub-networks for efficiency.
- [Transformers](https://kapyn.app/learn/transformers): The neural-network architecture, built on attention, behind nearly all modern LLMs.
- [Fine-Tuning](https://kapyn.app/learn/fine-tuning): Further training a pretrained model on a narrower dataset to specialise its behaviour.
- [Prompt Engineering](https://kapyn.app/learn/prompt-engineering): The practice of crafting model inputs to reliably get the outputs you want.
- [Diffusion Models](https://kapyn.app/learn/diffusion-models): Generative models that create images or audio by iteratively denoising random noise.
- [Quantization](https://kapyn.app/learn/quantization): Compressing a model's weights to lower precision so it runs faster and on less memory.
- [Chain of Thought](https://kapyn.app/learn/chain-of-thought): Prompting a model to reason step by step, which improves accuracy on complex tasks.
- [Context Window](https://kapyn.app/learn/context-window): The maximum amount of text a model can consider at once when generating a response.
- [Inference](https://kapyn.app/learn/inference): Running a trained model to produce outputs — the compute cost users actually pay at runtime.
- [RLHF](https://kapyn.app/learn/rlhf): Reinforcement Learning from Human Feedback — tuning models to match human preferences.
- [Embeddings](https://kapyn.app/learn/embeddings): Numeric vector representations of text or data that let machines compare meaning by distance.
- [Vector Database](https://kapyn.app/learn/vector-database): A database built to store and search embeddings by similarity, the backbone of RAG systems.
- [India](https://kapyn.app/learn/india)
- [Python](https://kapyn.app/learn/python)
- [Agentic AI](https://kapyn.app/learn/agentic-ai)
- [AI safety](https://kapyn.app/learn/ai-safety)
- [macOS](https://kapyn.app/learn/macos)
- [Rust](https://kapyn.app/learn/rust)
- [GPU](https://kapyn.app/learn/gpu)
- [AI Governance](https://kapyn.app/learn/ai-governance)
- [Sam Altman](https://kapyn.app/learn/sam-altman)
- [Reinforcement Learning](https://kapyn.app/learn/reinforcement-learning)
- [robotics](https://kapyn.app/learn/robotics)
- [multimodal AI](https://kapyn.app/learn/multimodal-ai)
- [Physical AI](https://kapyn.app/learn/physical-ai)
- [JavaScript](https://kapyn.app/learn/javascript)
- [Edge AI](https://kapyn.app/learn/edge-ai)
- [AI infrastructure](https://kapyn.app/learn/ai-infrastructure)
- [autonomous agents](https://kapyn.app/learn/autonomous-agents)
- [EU AI Act](https://kapyn.app/learn/eu-ai-act)
- [AI Regulation](https://kapyn.app/learn/ai-regulation)
- [Dario Amodei](https://kapyn.app/learn/dario-amodei)
- [coding-agents](https://kapyn.app/learn/coding-agents)
- [open-source](https://kapyn.app/learn/open-source)
- [China](https://kapyn.app/learn/china)
- [AI alignment](https://kapyn.app/learn/ai-alignment)
- [GPUs](https://kapyn.app/learn/gpus)
- [embodied AI](https://kapyn.app/learn/embodied-ai)
- [sovereign AI](https://kapyn.app/learn/sovereign-ai)
- [data centers](https://kapyn.app/learn/data-centers)
- [multi-agent systems](https://kapyn.app/learn/multi-agent-systems)
- [open-weight models](https://kapyn.app/learn/open-weight-models)
- [prompt injection](https://kapyn.app/learn/prompt-injection)
- [AGI](https://kapyn.app/learn/agi)
- [alignment](https://kapyn.app/learn/alignment)
- [Jalapeño](https://kapyn.app/learn/jalapeno)
- [open models](https://kapyn.app/learn/open-models)
- [open source AI](https://kapyn.app/learn/open-source-ai)
- [Gemini API](https://kapyn.app/learn/gemini-api)
- [WebAssembly](https://kapyn.app/learn/webassembly)
- [Recursive Self-Improvement](https://kapyn.app/learn/recursive-self-improvement)
- [data center](https://kapyn.app/learn/data-center)
- [Elon Musk](https://kapyn.app/learn/elon-musk)
- [quantum computing](https://kapyn.app/learn/quantum-computing)
- [voice AI](https://kapyn.app/learn/voice-ai)
- [semiconductors](https://kapyn.app/learn/semiconductors)
- [AI adoption](https://kapyn.app/learn/ai-adoption)
- [enterprise AI](https://kapyn.app/learn/enterprise-ai)
- [Agent Skills](https://kapyn.app/learn/agent-skills)
- [Agentic Workflows](https://kapyn.app/learn/agentic-workflows)
- [LLM agents](https://kapyn.app/learn/llm-agents)
- [React](https://kapyn.app/learn/react)
- [world models](https://kapyn.app/learn/world-models)
- [multimodal](https://kapyn.app/learn/multimodal)
- [iOS](https://kapyn.app/learn/ios)
- [Apple Intelligence](https://kapyn.app/learn/apple-intelligence)
- [Responsible AI](https://kapyn.app/learn/responsible-ai)
- [OpenTelemetry](https://kapyn.app/learn/opentelemetry)
- [Apache 2.0](https://kapyn.app/learn/apache-2-0)
- [Jensen Huang](https://kapyn.app/learn/jensen-huang)
- [Markdown](https://kapyn.app/learn/markdown)
- [AI](https://kapyn.app/learn/ai)
- [AI inference](https://kapyn.app/learn/ai-inference)
- [AI data centers](https://kapyn.app/learn/ai-data-centers)
- [Windows](https://kapyn.app/learn/windows)
- [AI coding agents](https://kapyn.app/learn/ai-coding-agents)
- [Amazon SageMaker HyperPod](https://kapyn.app/learn/amazon-sagemaker-hyperpod)
- [Apple Silicon](https://kapyn.app/learn/apple-silicon)
- [GRPO](https://kapyn.app/learn/grpo)
- [AI safety frameworks](https://kapyn.app/learn/ai-safety-frameworks)
- [MIT license](https://kapyn.app/learn/mit-license)
- [AI policy](https://kapyn.app/learn/ai-policy)
- [Superintelligence](https://kapyn.app/learn/superintelligence)
- [open-source AI models](https://kapyn.app/learn/open-source-ai-models)
- [computer vision](https://kapyn.app/learn/computer-vision)
- [open-weights](https://kapyn.app/learn/open-weights)
- [deepfakes](https://kapyn.app/learn/deepfakes)
- [AI security](https://kapyn.app/learn/ai-security)
- [MMLU](https://kapyn.app/learn/mmlu)
- [Linux](https://kapyn.app/learn/linux)
- [GPU clusters](https://kapyn.app/learn/gpu-clusters)
- [drug discovery](https://kapyn.app/learn/drug-discovery)
- [arXiv](https://kapyn.app/learn/arxiv)
- [distillation](https://kapyn.app/learn/distillation)
- [Humanoid robotics](https://kapyn.app/learn/humanoid-robotics)
- [frontier models](https://kapyn.app/learn/frontier-models)
- [agentic systems](https://kapyn.app/learn/agentic-systems)
- [Trump](https://kapyn.app/learn/trump)
- [existential risk](https://kapyn.app/learn/existential-risk)
- [GPU acceleration](https://kapyn.app/learn/gpu-acceleration)
- [ARC-AGI-3](https://kapyn.app/learn/arc-agi-3)
- [EU](https://kapyn.app/learn/eu)
- [IPO](https://kapyn.app/learn/ipo)
- [multi-agent](https://kapyn.app/learn/multi-agent)
- [benchmarks](https://kapyn.app/learn/benchmarks)
- [open-weight AI](https://kapyn.app/learn/open-weight-ai)
- [C#](https://kapyn.app/learn/c)
- [RSI](https://kapyn.app/learn/rsi)
- [benchmarking](https://kapyn.app/learn/benchmarking)
- [AI ethics](https://kapyn.app/learn/ai-ethics)
- [vibe coding](https://kapyn.app/learn/vibe-coding)
- [MCP server](https://kapyn.app/learn/mcp-server)
- [genomics](https://kapyn.app/learn/genomics)
- [knowledge graph](https://kapyn.app/learn/knowledge-graph)
- [ASR](https://kapyn.app/learn/asr)
- [post-quantum cryptography](https://kapyn.app/learn/post-quantum-cryptography)
- [LLMs (large language models)](https://kapyn.app/learn/llms-large-language-models)
- [scaling laws](https://kapyn.app/learn/scaling-laws)
- [human in the loop](https://kapyn.app/learn/human-in-the-loop)
- [DRAM](https://kapyn.app/learn/dram)
- [enterprise AI adoption](https://kapyn.app/learn/enterprise-ai-adoption)
- [HPC](https://kapyn.app/learn/hpc)
- [governance frameworks](https://kapyn.app/learn/governance-frameworks)
- [red teaming](https://kapyn.app/learn/red-teaming)
- [automation](https://kapyn.app/learn/automation)
- [High-Performance Computing](https://kapyn.app/learn/high-performance-computing)
- [alignment risks](https://kapyn.app/learn/alignment-risks)
- [autonomous systems](https://kapyn.app/learn/autonomous-systems)
- [edge computing](https://kapyn.app/learn/edge-computing)
- [protein folding](https://kapyn.app/learn/protein-folding)
- [quantum error correction](https://kapyn.app/learn/quantum-error-correction)
- [Genesis Mission](https://kapyn.app/learn/genesis-mission)
- [LoRA](https://kapyn.app/learn/lora)
- [Automatic Speech Recognition](https://kapyn.app/learn/automatic-speech-recognition)
- [llmops](https://kapyn.app/learn/llmops)
- [MLX](https://kapyn.app/learn/mlx)
- [supervised fine-tuning](https://kapyn.app/learn/supervised-fine-tuning)
- [Search](https://kapyn.app/learn/search)
- [sparse attention](https://kapyn.app/learn/sparse-attention)
- [prompt caching](https://kapyn.app/learn/prompt-caching)
- [GraphRAG](https://kapyn.app/learn/graphrag)
- [cybersecurity](https://kapyn.app/learn/cybersecurity)
- [speech-to-text](https://kapyn.app/learn/speech-to-text)
- [multilingual AI](https://kapyn.app/learn/multilingual-ai)
- [LLM inference](https://kapyn.app/learn/llm-inference)
- [text-to-speech](https://kapyn.app/learn/text-to-speech)
- [5G](https://kapyn.app/learn/5g)
- [tool use](https://kapyn.app/learn/tool-use)
- [AI chips](https://kapyn.app/learn/ai-chips)
- [LLM safety](https://kapyn.app/learn/llm-safety)
- [India AI Mission](https://kapyn.app/learn/india-ai-mission)
- [Model alignment](https://kapyn.app/learn/model-alignment)
- [on-device AI](https://kapyn.app/learn/on-device-ai)
- [AlphaFold](https://kapyn.app/learn/alphafold)
- [Nvidia GPUs](https://kapyn.app/learn/nvidia-gpus)
- [Nandan Nilekani](https://kapyn.app/learn/nandan-nilekani)
- [OCR](https://kapyn.app/learn/ocr)
- [Pull Requests](https://kapyn.app/learn/pull-requests)
- [recommendation systems](https://kapyn.app/learn/recommendation-systems)
- [AI Safety Institute](https://kapyn.app/learn/ai-safety-institute)
- [Multimodal reasoning](https://kapyn.app/learn/multimodal-reasoning)
- [formal verification](https://kapyn.app/learn/formal-verification)
- [CPU](https://kapyn.app/learn/cpu)
- [Greg Brockman](https://kapyn.app/learn/greg-brockman)
- [cryptography](https://kapyn.app/learn/cryptography)
- [AI-assisted coding](https://kapyn.app/learn/ai-assisted-coding)
- [South Korea](https://kapyn.app/learn/south-korea)
- [vendor lock-in](https://kapyn.app/learn/vendor-lock-in)
- [AI governance frameworks](https://kapyn.app/learn/ai-governance-frameworks)
- [UK](https://kapyn.app/learn/uk)
- [AI systems](https://kapyn.app/learn/ai-systems)
- [inference optimization](https://kapyn.app/learn/inference-optimization)
- [autonomous driving](https://kapyn.app/learn/autonomous-driving)
- [agentic workflow](https://kapyn.app/learn/agentic-workflow)
- [AI workloads](https://kapyn.app/learn/ai-workloads)
- [Xi Jinping](https://kapyn.app/learn/xi-jinping)
- [hallucinations](https://kapyn.app/learn/hallucinations)
- [secret scanning](https://kapyn.app/learn/secret-scanning)
- [Fields Medal](https://kapyn.app/learn/fields-medal)
- [SQL](https://kapyn.app/learn/sql)
- [AGIEval](https://kapyn.app/learn/agieval)
- [cloud computing](https://kapyn.app/learn/cloud-computing)
- [AI as a Service](https://kapyn.app/learn/ai-as-a-service)
- [coding agent](https://kapyn.app/learn/coding-agent)
- [synthetic data](https://kapyn.app/learn/synthetic-data)
- [drones](https://kapyn.app/learn/drones)
- [skills](https://kapyn.app/learn/skills)
- [Managed Agents](https://kapyn.app/learn/managed-agents)
- [flow matching](https://kapyn.app/learn/flow-matching)
- [robotic manipulation](https://kapyn.app/learn/robotic-manipulation)
- [LLM evaluation](https://kapyn.app/learn/llm-evaluation)
- [datacenter](https://kapyn.app/learn/datacenter)
- [synthetic data generation](https://kapyn.app/learn/synthetic-data-generation)
- [alignment research](https://kapyn.app/learn/alignment-research)
- [transparency](https://kapyn.app/learn/transparency)
- [reproducibility](https://kapyn.app/learn/reproducibility)
- [Digital India](https://kapyn.app/learn/digital-india)

## Knowledge catalog (Open Knowledge Format)

Structured, agent-readable bundles in text/markdown with YAML frontmatter (OKF v0.1). Updated hourly.

- [Catalog index](https://kapyn.app/okf): bundle index, story counts, tool inventory, category breakdown
- [Today's stories](https://kapyn.app/okf/stories): full 48h news digest grouped by category
- [AI tools radar](https://kapyn.app/okf/tools): active AI tools, models, and entities tracked by Kapyn