The house lead
Claude Opus 5
claude-opus-5
1M-token context, 128K max output, adaptive thinking
Impressive Claude demos with receipts
MagicClaude
A theatrical gallery of Claude prompts that look like magic because the setup, reveal, and verification are all visible. Every demo names the model it runs on and links back to official Anthropic sources.
Cast as of September 2026: Claude Opus 5 for most demos, Claude Fable 5.1 for long-horizon agent runs, Claude Sonnet 5 and Claude Haiku 4.5 for rehearsal.
Tonight’s show
One wow-demo, with the wires labeled
The Stage spotlights the one-prompt artifact prototype on Claude Opus 5: copy the prompt, check the house cues, and verify the numbers. The rest of the cabinet waits until the first reveal has landed.
Start with the stage pieces
Three demos that make a room lean forward
These are not benchmark claims. They are scripts for showing what a current Claude surface can do when the prompt gives it a clear role, source boundary, output shape, and verification loop.
Artifacts
Artifact Prototype
15+ lines
Artifact threshold signal
Give Claude a messy product idea and ask for a shareable interactive prototype with states, empty cases, and iteration notes.
Opus 5Beginner10-20 minutes
Documents
PDF Evidence Table
Text + charts
PDF analysis surface
Upload a PDF and ask Claude to separate claims, supporting pages, uncertainty, and follow-up checks into a reusable evidence table.
Opus 5Intermediate15-30 minutes
Vision
Vision UI Audit
Image input
Visual reasoning setup
Upload a screenshot and ask Claude to return prioritized UX, accessibility, and conversion issues with exact visual evidence.
Opus 5Beginner8-15 minutes
Tonight's cast
Every demo names its model
A prompt without a model name stops being reproducible the moment the lineup changes. These four are generally available as of September 2026. For most demos the cast is Claude Opus 5; the exceptions are called out on the demo pages.
The long-horizon headliner
Claude Fable 5.1
claude-fable-5-1
1M-token context, 128K max output, always-on adaptive thinking
The understudy who never misses a cue
Claude Sonnet 5
claude-sonnet-5
1M-token context, 128K max output
The stagehand
Claude Haiku 4.5
claude-haiku-4-5
200K-token context, 64K max output
Free tools
Find, remix, and rate the reveal
Prompt gallery
Demo Gallery
Filter impressive Claude demo prompts by format, difficulty, audience, and reveal type, then copy the exact prompt and verification checklist.
FreeClient-sideNo signup
Prompt generator
Prompt Remixer
Remix a Claude demo prompt by changing theme, output style, audience, difficulty, and constraints while preserving a verifiable structure.
FreeClient-sideNo signup
Demo checklist
Wow Meter
Score your Claude demo result for first-try reliability, faithfulness, surprise, usefulness, safety, and share readiness.
FreeClient-sideNo signup
Why this exists
The internet has prompt lists. It needs demo scripts.
A prompt list is easy to copy and hard to trust. A demo script says what to upload, what to ask, what Claude should reveal, and how to check whether the reveal is real.
MagicClaude stays in one lane: impressive Claude demos with reproducible prompts. For Claude news, useClaude Weekly. For scored model comparisons and pricing analysis, useClaude Reports. For deeper Claude Code engineering practice, useClaude Ships Code. For context-engineering theory, useClaude Context.
That boundary matters most around models. Naming Claude Fable 5.1 or Claude Opus 5 in a demo is reproducibility. Ranking them is a different job, done with methodology this site does not have.
Full cabinet
All reproducible demos
Data
Data Visualizer
Sandbox
Code-assisted analysis
Give Claude a small dataset and ask it to produce an interactive chart, assumptions, anomalies, and a shareable artifact version.
Opus 5Intermediate15-25 minutes
Claude Code
Codebase Debugging
Repo-aware
Agentic coding surface
Use Claude Code to read a repo, reproduce a bug, make the narrow fix, and report the tests it ran.
Opus 5 or Fable 5.1Advanced20-45 minutes
Computer Use
Browser Chore
Preview
Permissioned desktop workflow
Show Claude using browser/computer workflows on non-sensitive tasks while keeping permissions and prompt-injection risk visible.
Opus 5Advanced20-40 minutes
Research
Research Brief
Current
Cited web grounding
Ask Claude to research a current question, use multiple sources, and separate established facts from open questions.
Opus 5Intermediate15-35 minutes
Long Context
Long Context
1M tokens
Context window on current models
Give Claude a large transcript or document bundle and ask for a structured synthesis with decision logs, contradictions, and missing evidence.
Fable 5.1 or Opus 5Intermediate20-45 minutes
Agentic
Long-Horizon Agent
Fable 5.1
Long-horizon cast
Give Claude Fable 5.1 a task too long for one turn and ask it to plan, execute, checkpoint its own progress, and report what it verified.
Fable 5.1Advanced45-120 minutes
Model behavior
Effort Dial
Adaptive
Thinking effort, always on
Run one ambiguous task at low and high thinking effort, then read the two answers side by side to show what deliberation actually buys.
Fable 5.1 and Opus 5Advanced20-40 minutes
Primary-source sample
What the sources actually support
Claude Help Center
What are artifacts and how do I use them?
Artifacts are substantial standalone outputs such as documents, code snippets, HTML sites, SVGs, diagrams, and React components.
Claude Platform Docs
Claude Fable 5.1 overview
Claude Fable 5.1 (`claude-fable-5-1`) has a 1M-token context window, 128K max output, always-on adaptive thinking, and a default effort level of high in Claude Code and medium on Claude.ai and Cowork.
Claude Code Docs
Claude Code overview
Claude Code is an agentic coding tool available in terminal, IDE, desktop app, and browser surfaces.