AI Models & Assistants
Claude is our core — but the same skills travel. This section widens the lens to the whole AI world: the major assistants, how they differ, when to use which, and the techniques that transfer across all of them.
- Pick the right model for a job without re-learning the field every time
- Move between assistants — ChatGPT, Gemini, Grok, open models — without losing your technique
- Run capable models on your own machine, and know when that is worth it
- Combine Claude with local models and agent frameworks
Start here
If you read one page in this section, read this one. Everything below is a branch off it.
- How to choose a model — a durable framework that outlives any particular release.
Then follow whichever thread matches what you are doing.
The current flagships
The two September 2026 releases at the top of both stacks, read side by side.
- Claude Fable 5.1: what changed and how to migrate — three breaking changes, five betas, cache reads at a quarter of the price
- GPT-6 Astra: the field guide for Claude users — same list price, different bill, Critical-tier cyber classification
Coming from another assistant
You already know one tool and want your habits to carry over.
- ChatGPT for Claude users
- Gemini for Claude users
- Grok for Claude users
- Porting prompts across models — what transfers, what breaks
Comparing before you commit
- Claude vs GPT vs Gemini for coding
- Reasoning models compared
- What AI costs across providers
- Generative media: image, audio, video
Running models yourself
Local models trade some capability for privacy, cost control, and working offline. These pages cover when that trade is worth making.
- Run models locally with Ollama — the practical starting point
- Why a 35 KB system prompt breaks on a local model — context budgets, Ollama's silent middle-of-prompt truncation, and the failure signals to watch
- DeepSeek, Qwen & the open models
- DeepSeek V4-Flash-Vision: first native-vision frontier open weight
- Colibrì: run a 744B MoE from your SSD — expert streaming on 16–32 GB machines, and what it really costs in tok/s
- A private local AI stack
- Claude plus local models — using both, deliberately
Agents beyond Claude
- Local AI agents
- Open-source agent frameworks
- Claude, MCP and local tools
- A2A: the agent-to-agent protocol
- Agent payments: x402, MPP and AgentCore GA — how agents actually pay for things now
- OpenAI Agents API vs Claude Managed Agents — the two hosted agent loops side by side: permissions, vaults, budgets, network policy, sandbox cost
Everything in this section
The full list, including pages added since this index was written.
AI Models & Assistants
Claude is the core here, but the same skills travel. Start with how to choose, then go as deep as you need on any assistant, on running models locally, or on agents.
AI 模型全景：在 Claude、ChatGPT、Gemini 与开源模型之间做选择
一套经得起时间考验的框架，用于为任务挑选合适的 AI 模型——涵盖 Claude、ChatGPT、Gemini、Llama、Mistral、Grok 以及开源/本地模型——并说明哪些技能可以在它们之间通用。
Claude Fable 5.1: What Changed and How to Migrate
Anthropic shipped Claude Fable 5.1 and Mythos 5.1 on September 1, 2026: same $10/$50 price, cache reads cut to $0.25, three breaking API changes (forced tool_choice errors, one-way thinking blocks, append-only history) and five additive betas. The migration field guide for anyone pinning claude-fable-5 or claude-opus-5.
Claude Fable 5 & Mythos 5: The Flagship Field Guide
Fable 5 is Anthropic's most capable model. Its API refuses in-band (HTTP 200, stop_reason: refusal), has adaptive thinking only, and never returns raw thinking. If you're pinning a top-tier model, this changes how you write integrations. When to use it vs Opus 4.8, the fallback pattern, the prompting shifts.
Claude Opus 5: The Field Guide
Opus 5 shipped 24 July 2026 as Anthropic's fourth model in two months — same $5/$25 pricing as Opus 4.8 but more than double the Frontier-Bench score, 3× ARC-AGI-3 vs the next-best model, and a new xhigh/max effort tier. The benchmarks that matter, the two 400-errors that break naive migrations, and when Opus 5 replaces Fable 5 in your stack.
Claude Sonnet 5: The Field Guide
Sonnet 5 is Claude Code's new default and Sonnet 4.6's drop-in successor — but three API constraints will 400 you on migration and a new tokenizer produces ~30% more tokens for the same text. The pricing math, the effort re-mapping (Sonnet 5 medium ≈ Sonnet 4.6 high), and the prompting shifts that catch teams out.
面向 Claude 用户的 ChatGPT 指南
已经熟练使用 Claude，却需要用 ChatGPT？看看真正重要的差异、哪些经验可以迁移，以及何时该选哪一个。
GPT-5.6 & ChatGPT Work for Claude Users
OpenAI shipped the GPT-5.6 family (Sol, Terra, Luna) and ChatGPT Work on 9 July 2026. What actually changed for a Claude user — 1.05M context, a $1 tier, an agent that runs for hours, and the parts of the launch nobody is writing about.
GPT-5.6 August 2026 Update: Effort Slider, Free Think Button & the 272K Price Cliff
On 6 August 2026 OpenAI shipped a mid-cycle update to GPT-5.6 that most write-ups reduced to 'Sol got smarter.' The real story: a six-tier effort control that changes how you price a request, a free-tier Think button, a 272K token pricing cliff that re-prices the whole call, and a Fast mode that only runs on Sol. What each means for a Claude user.
GPT-6 Astra: The Field Guide for Claude Users
OpenAI's GPT-6 Astra went GA on September 4, 2026: 1.05M context, $10/$50 per MTok, a Critical cyber classification, and a 'recurrent depth' architecture that hides part of its reasoning. What's confirmed, how it stacks against Claude Fable 5.1 and Opus 5, the four API gotchas, and when a Claude user should reach for it.
Codex Agent Plugins & Catalog Federation (v0.147.0): The Practical Guide
On 7 August 2026 OpenAI shipped Codex CLI v0.147.0 — the release that turns Skills into portable Agent Plugins with a four-tier catalog, ships an --approve-for-me flag that doesn't do what it sounds like, quietly removes --full-auto, and opts you into MCP 2026-07-28. What every part actually does, in Claude-user terms.
Sign in with ChatGPT
OpenAI's new identity layer, three days old. What data actually crosses the line, what enterprise admins can do about it, and how to accept it in your app without giving OpenAI a foothold you didn't intend.
用 Ollama 在本地运行 AI 模型
用 Ollama 在你自己的机器上运行开放权重模型（Llama、Mistral、Qwen……）——私密、离线、运行免费。安装、CLI，以及如何从代码中调用它。
Claude vs GPT vs Gemini 编程对比
一个用于挑选前沿编程模型的框架（不是排行榜）——真正重要的因素，以及为什么在你自己的代码库上做一次评测胜过任何基准测试。
编程 Agent CLI 横向对比
Claude Code、Codex CLI、Gemini CLI 以及开源终端 agent —— harness（而不只是模型）如何决定你的结果，以及如何用 AGENTS.md 保持配置的可移植性。
Supabase Evals — Real-Backend Agent Benchmark
Supabase open-sourced a benchmark that runs Claude Code, Codex and OpenCode against real containerized Supabase stacks — not mocks. Here's what it measures, what it found about how models actually use docs and skills, and how to run it locally in an evening.
Apple Foundation Models 3 & the On-Device LLM Stack for Claude Users
Apple's third-generation Foundation Models (AFM 3) landed at WWDC 2026 with a 3B dense on-device model, a 20B sparse MoE that runs on iPhone, and a Swift-native framework that now lets you bring your own LLM provider — including Claude. What actually shipped, the non-obvious mechanisms, and how it changes the on-device story for anyone who ships Claude apps.
给 Claude 用户的 Gemini 指南
熟悉 Claude，却需要用 Google 的 Gemini？重要的差异——Gems、Workspace 集成、超大上下文、多模态——以及哪些经验可以直接迁移。
Gemini 3.6 Flash, Flash-Lite & Flash Cyber for Claude Users
On 21 July 2026 Google skipped Gemini 3.5 Pro entirely and shipped three Flash-tier models instead. Real prices, real benchmarks, the migration gotchas, and the parts most write-ups miss — including a cybersecurity fine-tune that beat Claude Opus 4.6 at V8 bug-hunting.
Gemini 3.7 Flash for Claude Users: Coding-Agent Workhorse at Half Price (Aug 2026)
On 13 August 2026, three weeks after 3.6 Flash, Google shipped Gemini 3.7 Flash at half the introductory price with a 16-point DeepSWE jump. The gotchas most write-ups skip: an expiring intro price, a 3× throughput advantage that matters for agent loops but not chat, and where it still trails Claude Sonnet 5.
写给 Claude 用户的 Grok 指南
已经用惯了 Claude，却需要用 xAI 的 Grok？看看真正重要的差异——实时 X/网页搜索、兼容 OpenAI 的 API，以及 Grok Build 编码智能体——外加所有可以直接迁移的东西。
Muse Code + Muse Spark 1.2: Meta's Terminal Coding Agent
On 5 August 2026 Meta shipped Muse Code, a terminal coding agent, and Muse Spark 1.2, the model co-trained with it. What the append-only event log actually buys you, why the contributor tier is 12.5× cheaper than standard, how it stacks up against Claude Code on Terminal-Bench 2.1, and the three bundled skills that are worth stealing regardless of which agent you use.
开源 AI Agent 框架
一份中立于厂商的地图，梳理构建 LLM agent 的主要开源框架——LangGraph、LlamaIndex、AutoGen、CrewAI 以及极简循环方法——以及如何选择。
Native Multi-Agent APIs: OpenAI's Responses Multi-Agent Beta vs Building It Yourself
On 9 July 2026 OpenAI shipped the Responses API multi-agent beta and Sol Ultra Mode — the first frontier vendor to make 'the model spawns and synthesises N subagents in one HTTP request' a first-class primitive. What actually ships, the API shape, the cost math, what Anthropic does instead, and when the native primitive beats a DIY fan-out.
A2A: The Agent-to-Agent Protocol
MCP connects agents to tools. A2A connects agents to other agents — across frameworks, clouds, and companies. Learn the Agent Card, tasks, the eight lifecycle states, streaming vs. webhooks, and how A2A composes with MCP.
Agent Payments: x402, MPP and AgentCore GA
How AI agents actually pay for things now — the x402 HTTP-402 protocol, the exact/upto/batch-settlement schemes, facilitators, Amazon Bedrock AgentCore Payments' Aug 2026 GA (payment sessions, Coinbase Bazar MCP, Stripe Privy, Machine Payment Protocol), and how to add a paywall or a spending agent without inventing a wallet.
跨模型迁移提示词
在 Claude、GPT、Gemini 与开源模型之间迁移一段提示词——哪些能干净地转移，哪些需要按模型调整，以及一套迁移工作流。
Model Routing Patterns: Cascades, Classifiers, and What Actually Ships
The design patterns for sending each request to the right AI model — rule-based, classifier, complexity, cascade, ensemble, fallback. When to use each, what breaks, and the recipes that actually ship in 2026.
DeepSeek、Qwen 与开放权重浪潮
强大的开放权重模型（DeepSeek、Qwen、Llama、Mistral）已经缩小了大部分差距。“开放权重”意味着什么、相对于闭源前沿模型的取舍，以及如何使用它们。
Qwen3.8-Max: The First Open-Weight Max-Class Model (When the Weights Ship)
August 3, 2026: Alibaba's Qwen3.8-Max shipped API-only with a 2.4T parameter MoE (95B active), a 1M context, dual OpenAI/Anthropic-spec APIs, and benchmark scores that beat GPT-5.6 Sol Max on OSWorld. What's genuinely new, what 'open weight' actually means when the Hugging Face page is empty, and the sparsity math that makes self-hosting expensive.
DeepSeek V4-Flash-Vision: First Native-Vision Frontier Open Weight
On Aug 31 2026 DeepSeek open-sourced V4-Flash-Vision-Exp — a 305B MoE that adds native vision to V4-Flash without hurting text agent scores, matches or beats Claude Opus 4.8 on ApexBench / Agents' Last Exam / ZeroBench at a fraction of the price, and ships MIT. The specific numbers, the hard 384-token-per-image cap most guides ignore, and the concrete way to run it.
面向 Claude 用户的 Kimi K2
熟悉 Claude，又对 Moonshot 的 Kimi K2 感到好奇？这是一款你真的可以让 Claude Code 直接对接的开放权重智能体模型——因为 Moonshot 提供了 Anthropic 兼容的端点。哪些是真正的不同，哪些又可以照搬。
GLM-5.2: Open-Weight Frontier Coding Model
Z.ai's 753B MoE (~40B active), 1M-token context, MIT-licensed — closes to within a point of Claude Opus 4.8 on long-horizon coding, at roughly one-sixth the API cost. What IndexShare actually does, how to run it locally, and where it beats Claude Code in the wild.
Inkling: Thinking Machines' Open-Weights Model
Mira Murati's Thinking Machines Lab shipped Inkling — a 975B-parameter Apache 2.0 multimodal MoE with a continuous thinking-effort dial. What the launch coverage got wrong, why it deliberately loses benchmarks, and what the effort knob actually does.
Kimi K3: World's Largest Open-Weight Model
Moonshot AI's 2.8-trillion-parameter open-weight model beats Claude Opus 4.8 on many benchmarks and tops the Frontend Code Arena. What makes it architecturally different, when the cache-hit pricing flips the cost math, and whether it's worth reaching for.
Running Kimi K3 Locally: vLLM, DSpark & the Real Hardware Bill
The definitive practical guide to self-hosting Moonshot's 2.8T open-weight Kimi K3 in 2026. What DSpark speculative decoding actually does, minimum hardware (1× DGX B300 or 16× DGX Spark), the exact vLLM serve commands, the prefill/decode asymmetry nobody warns you about, and when hosted inference undercuts your own rack.
Inside Grok Build: Reading an Open Agent Harness
xAI published the full Rust source of its coding agent under Apache 2.0 — the agent loop, tool layer, TUI and extension system. What the crate map teaches you about building agents, why 'open source' here means source-available, and what the data-collector code still in the tree tells you.
Poolside Laguna: Open-Weight Coding Models That Punch Above Their Class
Poolside AI's Laguna family (XS 2.1, S 2.1, M.1) — MoE coding models with 3B–23B active params, 1M-token context, OpenMDW-1.1 license. What the 8B-active S 2.1 actually beats, why the 3B-active XS 2.1 runs on a MacBook, the first-known RL-in-FP8 training pipeline, and when to reach for it over Claude, GLM-5.2 or Kimi K3.
生成式媒体：图像、音频与视频 AI
超越文本对话——一张关于图像、视频、语音和音乐 AI 的地图：主要工具、可迁移的技能，以及不可忽视的权利与伦理问题。
MiniMax H3 (Hailuo 3.0): The Open-Weight Omni Video Model
MiniMax open-sourced a 33B single-stream transformer that generates 4–15s of 2K video with native stereo audio in one pass. The license quietly blocks self-hosting in the US/EU/UK/Korea, the open base is 768p (not 2K), and ~13B of 33B params are AdaLN branches you can skip at inference. What H3 actually is, what it costs, and when it beats the closed video APIs.
AI 到底要花多少钱（跨服务商对比）
一个跨服务商比较 AI 成本的持久框架——定价原型、如何估算一个工作负载、削减账单的杠杆，以及开源/本地何时更划算。
Why AI Agents Burn Tokens (and How to Cap the Bill)
Agentic runs cost 50–1000x a chat — not because the model is dumber, but because of the context snowball. The mechanism, the surprising numbers, and the four levers that actually cut the bill across Claude, GPT, Gemini, and open models.
构建本地 AI 智能体
完全在你自己机器上运行的自主智能体，使用本地开放权重模型——私密、离线、运行免费。架构、权衡取舍，以及如何上手。
Claude + 本地模型：混合模式
让 Claude 与本地开放权重模型协同工作——路由、先草稿后精修、隐私脱敏、批量预处理——用前沿模型的推理能力，换取本地模型的隐私与成本优势。
AI Gateways: LiteLLM, OpenRouter, Portkey, Vercel
A production AI gateway is the missing router between your app and every model — Claude, GPT, Gemini, Llama. Compare LiteLLM, OpenRouter, Portkey, and Vercel AI Gateway, then wire Claude Code through your own proxy.
本地编码智能体（以及它们如何与 Claude 搭配）
在你的机器上运行并编辑你仓库的编码智能体 —— Aider、Cline、Continue、OpenHands、Claude Code —— 用 Claude 追求质量，或用本地模型保护隐私。
用 MCP 把 Claude 连接到本地工具与智能体
Model Context Protocol 让 Claude 编排在你自己机器上私密运行的工具、数据与智能体——这就是「Claude 作大脑、本地提供能力」的协同。添加或自建一个本地 MCP 服务器。
搭建私有本地 AI 栈（端到端）
把各部分拼到一起：本地模型（Ollama）+ 模型无关的智能体 + 通过本地 MCP 服务器提供的工具 —— 一个运行在你自己机器上的私有助手，Claude 作为可选的智能层。
本地智能体还是 Claude？一份决策指南
完全本地的智能体、由 Claude 驱动，还是混合方案？决定这一选择的因素——隐私、任务难度、成本、延迟、可靠性——以及为什么混合方案往往胜出。
保护本地与混合智能体
能够编辑文件、运行命令的智能体既强大又危险。最小权限、沙箱、人工审批、提示注入防御、预算上限——如何安全地运行智能体。
推理模型对比
对比 Claude、GPT、Gemini、DeepSeek 和 Qwen 的思考预算与推理强度——何时该为思考花费 token，何时不该。
Full-Duplex Voice AI: Why Voice Agents Suddenly Got Real
GPT-Live and the new speech-native models listen and talk at the same time. What full-duplex changes, how it works, and the voice-AI landscape now.
Claude Voice with Opus & Sonnet (July 2026): Reasoning-First Voice vs GPT-Live
Anthropic opened Claude voice to Opus, Sonnet, and connected apps on July 23, 2026 — but kept the turn-based stack. What actually changed, when to switch models mid-call, and how the design bet differs from OpenAI's full-duplex GPT-Live.
Token Speed: Why AI Inference Suddenly Got 10-15× Faster
Wafer-scale chips and LPUs are serving frontier models at hundreds of tokens per second. What actually limits inference speed, the new hardware race, and why fast tokens change agents, voice and reasoning.
Computer-Use Agents Compared
Claude, GPT and Gemini can all drive a screen now — but each exposes a different action space, coordinate contract and safety gate. What actually breaks, and how to build a loop that survives it.
Shared-Login Browsers for AI Agents: The ego lite Pattern
A new class of browser — best exemplified by ego lite, which hit #1 on GitHub Trending on July 24, 2026 — lets Claude Code, Codex, Cursor, and any other coding agent drive an isolated Chromium session that inherits your real logins. What Spaces and semantic snapshots actually solve, why it undercuts Playwright for agent work, and the security cost of sharing session cookies with an autonomous process.
Cloudflare Kitesurf: The First Browser Runtime Built for AI Agents (Not Humans)
On August 6, 2026 Cloudflare launched Kitesurf, a stateless, agent-first browser that runs entirely in V8 isolates on Workers. It uses 3-7x less CPU and memory than Chromium on common agent tasks — at the cost of ~1.7x wall-clock time. Written in Rust, using Firefox's CSS parser (Stylo) and the Blitz rendering engine, drop-in compatible with Puppeteer/Playwright via the Chrome DevTools Protocol. This page: what changed, when to reach for it instead of headless Chrome, how to wire it into Claude Code / Codex / your MCP client, and the four things it flat-out can't do yet.
SKILL.md: The Cross-Agent Open Standard
One SKILL.md file, thirty-two agents. What the Agent Skills open standard actually is, how progressive disclosure keeps context cheap, what breaks portability, and how to write a skill that runs unchanged in Claude Code, Codex, Gemini CLI and Cursor.
Colibrì: Run a 744B MoE From Your SSD (and What It Really Costs You)
Colibrì is a pure-C, zero-dependency engine that runs GLM-5.2 (744B), Kimi K3 (2.8T), DeepSeek V4.1 Flash and six other frontier MoE models on a 16–32 GB machine by keeping the dense layers in RAM and streaming the routed experts from NVMe. This page explains the memory hierarchy, the numbers that decide whether it is usable for you (0.05 to 6.8 tok/s), the SSD-wear and speculative-decoding gotchas, and the exact commands to get a first token.
OpenAI Agents API vs Claude Managed Agents: The Hosted Agent Loop, Compared
On September 10, 2026 OpenAI put the Codex harness behind a single API call: the Agents API public beta — durable sessions, hosted or self-hosted sandboxes, automatic compaction, tool search, subagents. Anthropic has run the same shape since April as Managed Agents. Both use the same four primitives; they differ on permissions, secrets, budgets, network policy and what the sandbox costs. The side-by-side, the request shapes, the six gotchas, and when to pick which.
Why a 35 KB System Prompt Breaks on a Local Model
Moving a big Claude/GPT system prompt to Ollama or another local runtime: the context-budget math, Ollama's silent middle-of-prompt truncation, the failure signals of context exhaustion, and the restructuring that actually works.