#claude articles

Claude Desktop Extensions とは｜.mcpb で MCP を簡単導入

Anthropic が公開した Desktop Extensions(.mcpb)を解説。ファイルをダブルクリックするだけで MCP サーバーを Claude Desktop に追加でき、手動 JSON 編集やターミナル操作が不要に。mcpb init / mcpb pack での梱包手順、機密の扱い、企業配布、エンジニアの稼ぎへの効き方まで整理。

Claude Opus 4.5 の料金は？入力$5・出力$25・Batch割引を解説

Claude Opus 4.5 の API 料金は入力 $5・出力 $25/百万トークン。前世代 Opus 4.1 から 67% 値下げで常用圏へ。Batch API 50% 割引・プロンプトキャッシュ・Extended Thinking の追加コストまで整理。

Claude Opus 4.6 料金 — $5/$25・Fast mode・Batch 割引まとめ

Claude Opus 4.6 の API 料金は入力 $5・出力 $25/百万トークン。Opus 4.1 比 67% 値下げで常用圏へ。Fast mode（$30/$150）・Batch API（50% 割引）・1M コンテキスト標準料金・プロンプトキャッシュまで整理。

What Is Claude Sonnet 4.6 | Specs, Capabilities, and Differences from Opus

A breakdown of Claude Sonnet 4.6's official specs and features. Model ID is claude-sonnet-4-6, context window is 1M tokens, with support for both Extended Thinking and Adaptive Thinking. Covers comparisons with Opus 4.8 and Haiku 4.5, 300K output via Batch API, and how to use it on AWS Bedrock and Vertex AI — all from primary sources.

Claude 'Taking Longer Than Usual': Causes, Fixes, and Prevention

How to Use Claude on Windows 11 | Web, Desktop & Code: Which to Choose

Claude Tips | Boost Answer Quality with Extended Thinking, Projects, and Files

How to Use Claude Talk | Voice Conversation & /voice in Japanese

Claude Fast Mode | How to Use 2.5x Speed, Pricing & Subscription Details

A guide to Claude's Fast Mode for users who want faster output — boosting generation speed up to 2.5x. Covers supported models (Opus 4.8/4.7/4.6), how to use /fast in Claude Code and speed: fast in the API, pricing for Opus 4.8 at $10 input / $50 output per million tokens, the important note that even subscribers are billed directly to usage credits, and how OTPS (not TTFT) is what actually speeds up.

Claude Fable 5 Sudden Suspension | Migrating to Opus 4.8 and Risk Diversification Strategies

Can You Use Claude on 2 PCs? Sync, Cowork Limitations, and Setup Guide

Claude × Meta Ads MCP | Automate Ad Management with Natural Language Using 29 Tools

claude-opus-4-1 Deprecation | Migration Guide and temperature 400 Error Fix

How Much Does Claude Sonnet Cost | API and Pro/Max Monthly Plans Explained

Explains Claude Sonnet 4.6 pricing for both API and Pro/Max monthly subscriptions. API rates are $3 input / $15 output per million tokens with no extra charge beyond 200K context. Batch API offers 50% off, prompt caching cache reads drop to 10%, and both discounts can be stacked.

What Is Claude Opus 4.5 | 76% Token Reduction Compared to Sonnet 4.5 Explained

A breakdown of Claude Opus 4.5's capabilities and pricing. At $5 input / $25 output — Sonnet-level pricing — it delivers roughly 67% cost reduction over Opus 4.1, and at medium Effort achieves Sonnet 4.5-equivalent accuracy with 76% fewer tokens. Covers how to use the Effort parameter and how to leverage prompt caching and batch inference.

How to Use Claude via NuGet | Official Anthropic SDK for C#

A guide for C# / .NET developers who want to use Claude via NuGet. Covers the official Anthropic package, the dotnet add package command, minimal code, API key setup, streaming, IChatClient integration, the gotcha that versions 3.X and below were a completely different package, and how it differs from the community Anthropic.SDK — all based on official documentation.

Claude Haiku 4.5 Performance | Speed and Cost-Efficiency Verified with Benchmarks

We examine Claude Haiku 4.5's performance across three dimensions: coding accuracy on SWE-bench Verified at 73.3% — on par with the previous-generation Sonnet 4 — response speed 4–5x faster than Sonnet 4.5, and pricing at $1 input / $5 output per million tokens. We also cover the 200K context limit and guidelines for choosing between models.

Claude Download for PC | Quickest Steps for Windows & Mac Sign-In

For anyone wanting to download Claude on PC: this single page covers the quickest steps for Windows and Mac, system requirements, sign-in, adding Claude Code (CLI), and common stumbling points after download — all based on official Anthropic information.

Claude Does Not Support JCB | Payment Methods and V-Preca / Wise Alternatives

Claude (Pro/Max/Team/API) does not support the JCB card brand. This article covers the list of accepted card brands, the structural reason JCB is rejected, and alternative payment methods such as V-Preca and Wise for those who only have a JCB card.

Claude Error X Search | 3 Steps to Check Outages in Real Time

A 3-step guide to determining in 30 seconds whether a Claude error is a widespread outage or just your own issue. Search 'Claude outage' on X in the Latest tab — if posts from the last 10 minutes appear continuously, it's a widespread outage.

Recommended MCP Servers for Claude | Official, Popular, and How to Choose by Use Case

A breakdown of recommended MCP servers for Claude by use case. Covers the 7 official reference implementations (Filesystem, Git, Fetch, etc.) as a starting point, with guidance on choosing and setting up third-party options like GitHub, Notion, and Slack. Also explains the Free plan's 1-connection limit and permission scope considerations.

Claude Windows Path Issues | where claude and How to Set Up PATH

A one-page reference — based on official Anthropic documentation — for the 'path' issues you'll encounter when using Claude / Claude Code on Windows: where claude.exe actually lives, how to fix 'claude not found' errors when PATH isn't set, the OneDrive home folder problem, and how to write backslash paths in MCP config files.

Claude Windows MCP Setup | Steps for config.json and OneDrive Issues

A step-by-step guide to adding MCP servers to Claude Desktop on Windows, covering Node.js setup, PowerShell execution policy, how to write claude_desktop_config.json, OneDrive path problems, and fully quitting the app when changes don't take effect.

How to Install Claude on Windows | Choosing Between Desktop and Code

For anyone looking to set up Claude on Windows 10/11: a guide covering when to use the Desktop app versus Claude Code (CLI), the fastest installation methods via MSIX/PowerShell/WinGet, admin privileges and virtualization settings required for Cowork, and common pitfalls — all based on official Anthropic documentation.

How to Load PDFs in Claude | Exporting Conversations Explained

A complete guide to both loading PDFs into Claude and exporting conversations to PDF. Claude.ai supports drag-and-drop loading; the API offers three methods — URL, Base64, and Files API — supporting up to 32 MB and 600 pages. For exporting conversations to PDF, the official export feature and Chrome extensions are the two practical options.

How to Use Claude × Figma | Design to Code with Dev Mode MCP

A guide to generating code from designs by connecting Claude and Figma via Dev Mode MCP. Covers three integration paths — Remote MCP, Desktop MCP, and Code to Canvas — with required plans, setup steps, and three common pitfalls to know on day one.

How to Use Claude × Canva | Automated Design Creation with Connectors and MCP

Learn how to integrate Claude and Canva to automatically generate designs. Covers three connection methods — the official Canva Connector, direct MCP, and AI Connector — along with the background behind Anthropic's AI Connectors and Canva's official integration, and step-by-step instructions for calling Canva from the Claude chat interface.

What Is Claude Sonnet 3.5 | Performance of the Legacy Model and Differences from Sonnet 4.6

Claude Sonnet 3.5 was retired on October 28, 2025, and new API calls are no longer possible. This article covers the performance, features, and pricing of both versions (20240620 / 20241022) and explains the differences from the current Claude Sonnet 4.6 along with migration steps.

Claude Opus 4.1 / 4.5 / 4.6 / 4.7 | Performance & Pricing Comparison

A four-axis comparison of Claude Opus 4.1 / 4.5 / 4.7 across SWE-bench score, pricing, speed, and migration risk. For autonomous agents, Opus 4.7 with SWE-bench 87.6% is the top choice; for cost savings, Opus 4.5 at $5 input is optimal; to avoid breaking changes, sticking with Opus 4.6 is the practical solution.

What is Claude Opus 4.1 | SWE-bench, Pricing, and Changes Explained

A breakdown of Claude Opus 4.1's changes, pricing, and API identifier usage. At the same price as Opus 4 ($15 input / $75 output per 1M tokens), it enhances multi-file refactoring and long-document search quality, achieving 74.5% on SWE-bench Verified. Includes guidance on when to pin a snapshot for stable production behavior.