Anthropic's Claude assistant and agentic coding tool, with chat apps, Projects, skills, connectors and Claude Code.
Updated EU status checked Last shown
Sources: the vendor's own pages (linked above) and creator videos at the linked timestamps. EU status is information, not legal advice; see how we check.
Claude is Anthropic's AI assistant, offered on web, desktop, iOS and Android. It can search the web, create files, run code, keep memory across conversations, build Artifacts, and connect to external apps and tools. Paid plans add Projects, Claude Code, scheduled and handed-off tasks, Claude in Chrome, Microsoft 365 integration and access to additional Claude models.
Claude Code is the agentic coding and automation harness within the Claude product family. YouTube creators describe using Claude with custom skills, Projects, connectors and scheduled tasks for non-coding workflows, including scripting and headless video editing with FFmpeg, subagent setups, browser automation, and running other models on local files. Reviewers note it uses screenshots in a loop for visual understanding and can drive a browser and Blender.
Access is through a Free plan with limited usage, Pro, Max and Ultra individual plans with higher usage limits, and Team and Enterprise plans. Usage limits reset every 5 hours and weekly, and the speed at which they are reached depends on model, chat length and file size. Anthropic states that Claude is available globally with regional data residency options via AWS Bedrock, Google Vertex and Microsoft Foundry.
Watch the moment: the exact second top creators show it. Videos load only when you press play, via youtube-nocookie.com.
Used as one of Peter's three harnesses for his skills. Matt builds the chief of staff scheduled task, advisor skill and skill-editing loop in Claude, with connectors and cloud scheduling. Peter says Opus 5.5 is a game changer for video editing and motion graphics.
Harness used to run Fable 5.2 on local files and projects for most of the tests.
The creator used the Claude Code app with Fable 5.2 to generate the Megabonk-style game in about 16 minutes at medium reasoning.
Matt shows how he uses Claude projects with custom skills to write short-form scripts and fully edit talking-head A-roll (cuts, B-roll, subtitles, music) using FFmpeg headlessly. He builds it from scratch on a new account and iterates with feedback, updating skills each time.
Used as the engine behind all three workflows: video editing skills, the subagent council and the Instagram automation. It uses screenshots in a loop for visual understanding and can drive a browser and Blender.
Discussion of Anthropic's S-1 filing: Claude drives the company's recursive 'closing the loop' improvement flywheel, with over 90% of code contributions written by Claude and enterprise customers growing rapidly. The creator is critical of the governance structure and safety contradictions.
Harness used for most tests, allowing multiple projects, local file access, a max effort setting and an ultra code mode that spins up many agents.
Redesigned projects in Claude Code: set a goal and repo, and a coordinator routes work to threads, each a cloud session on its own branch. In beta for select Pro and Max users without existing projects.
Harness used for most demos, allowing multiple projects, parallel agents and local file access.
Used to build the V1 of The Librarian 2 from a long voice prompt. It ran about 1.5 hours, opened the game in its own browser, tested and rebalanced itself, and produced about 10,400 lines of code across 33 modules with no asset files.
Agentic harness used to run Opus 5 on local files, automatically launching pages in a browser and fixing bugs.
Can call Codex through its CLI. The creator built a skill where Fable plans, Codex GPT-5.5 writes and Claude verifies.
Anthropic's coding agent, called its cash cow, which Claude Tag evolves from. The creator describes it as the second UI paradigm, an app on your computer.
Great but used less by the creator because he runs out of quota quickly. Uses CLAUDE.md instead of agents.md and has cloud agents.
Shown configured to run GLM 5.2 via Z.ai's coding helper and settings.json model overrides, then used for the ray tracing test.
Creator used the Claude Code desktop app to run Fable 5 on agentic coding tests, noting it auto-verifies its work in the browser.
Claude Code's /loop command runs a prompt or command on a recurring interval, e.g. every 5 minutes comparing the build against spec.md until complete. Can also be run without an interval until a goal is reached.
Coding agent launched in research preview in February 2025. Claude-authored code share at Anthropic rose from low single digits to over 80%. The creator uses it for underspecified prompts and long-running tasks.
Closest neighbours by meaning, not just by category.
App · Anthropic
In-app browser in the Claude Code desktop app with element annotation for requesting edits to web pages.
Agent · Anthropic
Anthropic feature that lets Slack teams tag Claude as a teammate to take on tasks and work in the background.
Agent · Anthropic
Claude Code mode that fans out many parallel sub-agents for deep tasks such as full codebase reviews.
App · Anthropic
Claude Code permission mode where a classifier approves safe tool calls and blocks risky ones instead of prompting.
App · Anthropic
Anthropic merges Claude Cowork, chat and Design into one Claude app with docs, slides and scheduled background tasks.
App · Anthropic
Claude in Chrome side panel now runs a Claude Cowork session shared with Claude's desktop, web and mobile apps.
