Post-Cutoff

A visualization of every Claude effort level, made with Opus 5.5 and Sonnet 5.5

DreW (@devteamdrew) X 229,228 views as of 27 October 2026

Why it is here

~218k views; explains Claude’s effort settings visually.

Description

Description written by Gemini from the videoGemini 3.8 Flash, 5 October 2026

Summary
Created by DreW (@devteamdrew) and generated with Claude Opus 5.5 and Sonnet 5.5, this animation humorously visualizes the escalating computational “effort levels” available for Claude AI models. A friendly, terracotta-colored box-shaped robot mascot acts out each thinking tier, transitioning from casual typing to a transcendent rainbow levitation.

What is shown

  • [00:00 - 00:05] effort: low: The robot opens a laptop, taps a few keys, and quickly closes it.
  • [00:05 - 00:11] effort: medium: The robot dons a headset with a microphone and a miniature keyboard, while a circular loop of pseudocode (while (think), const plan = ..., if (test()) return "done") scrolls around its head.
  • [00:12 - 00:17] effort: high: The robot hunches intently over a chessboard, shuffling pieces at hyper-speed before celebrating.
  • [00:18 - 00:24] effort: xhigh: Equipped with round spectacles between two buzzing green Tesla coils and a boiling flask, the robot channels electricity with matrix-like data cascading in its eyes.
  • [00:25 - 00:33] effort: max: Heavy mecha armor snaps onto the robot with an energy HUD reading EFFORT: MAX (PWR 80% ramping to PWR 100%), culminating in an expanding shockwave ring.
  • [00:34 - 00:45] effort: ultracode: The robot’s eyes illuminate purple as it multiplies into an infinite perspective army of parallel clones marching in synchronization.
  • [00:46 - 00:58] effort: ultrathink: The character body shifts into radiant rainbow tones with crystalline sparkling eyes, levitating off the ground over pulsing concentric energy ripples.

Claims & numbers

  • HUD indicator displays PWR 80% increasing to PWR 100% during the max effort sequence [00:29 - 00:30].
  • No quantitative benchmark scores, speed multipliers, or prices are explicitly stated.

Notable quotes

  • (No spoken dialogue)
  • On-screen pseudocode loop: while (think) { for (const f of plan) ... if (test()) return "done"; } [00:08]

Assessment
This is a community-made creative showcase and visual metaphor illustrating model effort parameters and reasoning budgets. It is an artistic interpretation rather than an official benchmark evaluation or product release by Anthropic.

Lyrics & themes

  • Format: Fully instrumental, accompanied by retro cartoon foley and synthesized sound effects matching each tier’s activity (keyboard clatter, electrical hum, power-up drones, and celestial chimes).
  • Themes: Test-time compute scaling, agentic decomposition, parallel swarm execution, and the philosophical “enlightenment” of extended model reasoning.

Lore & references

  • Terracotta Mascot: Styled in Anthropic’s signature terracotta/warm peach aesthetic, symbolizing Claude.
  • Thinking / Effort Budget: References the user-configurable effort parameters across modern LLM interfaces.
  • ultracode (Swarm Grid): Depicts massive parallel agentic worktrees and self-coordinating coding sub-agents.
  • ultrathink (Cosmic Levitation): Alludes to deep reasoning and search-based deduction that transcends single-pass generation.

Visual style & craft
The video utilizes a 2D illustrated storybook aesthetic with textured parchment paper backgrounds, clean ink linework, and soft cross-hatching. The motion design combines smooth vector transformations, particle lighting, and crisp UI HUD elements, characteristic of code-driven animation workflows (such as browser canvas or Remotion pipelines) orchestrated by LLM-assisted code generation.

Described by gemini-3.8-flash on 2026-10-05 from the video’s audio and frames.

How AI was used to make it

AI made all of it of this video: visuals from ai-written code.

  • Code-rendered 2D motion graphics generated from Opus 5.5 and Sonnet 5.5

Tools named: Claude Opus 5.5, Claude Sonnet 5.5

Classified by gemini-3.8-flash from our description of the video (high confidence), 25 October 2026. How AI is used to make videos, across the whole log

Made by AI

How we know
X post (2026-09-30): ‘A visualization of every Claude effort level, made with Opus 5.5 and Sonnet 5.5.’
Human role
Concept by the creator; process not described.
Pipeline
Opus 5.5 + Sonnet 5.5 → code-rendered motion graphic (59 s)
Series
Code-rendered film (LLM writes the program that draws every frame)

Related

  1. Model releases 90 days after the cutoff

    Anthropic releases Claude Sonnet 5.5

  2. Model releases 84 days after the cutoff

    Anthropic releases Claude Opus 5.5

More videos on this

  1. YouTube

    “Gave Opus 5.5 donald’s prompt, Midjourney, and a moodboard. 12 hours later, woke up to this”

    anabology (@anabology)20,278,966 views as of 15 Nov

  2. X

    Western civilization in 2 minutes 16 seconds

    vittorio (@IterIntellectus)“i asked claude to make a video on western civiization”15,209,393 views as of 23 Oct

  3. X

    “The Clodyssey”

    anabology (@anabology)Claude, another 12 hours12,576,393 views as of 27 Oct

  4. X

    “Build what you think you look like”: Claude Sonnet 5.5 vs GPT-6 Astra

    Siya (@siyabuilt)7,259,229 views as of 27 Oct