# LLM context windows and token grids > A Portuguese-language explainer pairs a chat transcript with separate input and output token grids and an LLM probability panel, showing how successive questions and answers accumulate within a limited context. - Kind: Video - Page: https://meetreference.com/v/video-llm-context-windows-and-token-grids-b2a86d/ - Ready-to-use agent handoff: https://meetreference.com/v/video-llm-context-windows-and-token-grids-b2a86d/agent.txt - Exact prompt as plain text: https://meetreference.com/v/video-llm-context-windows-and-token-grids-b2a86d/prompt.txt - Authorship: AI-made, per the source - Made with: Not stated - Model: Opus 5.5 - Built with: not stated - Skills or plugins: JavaScript - Style: Explainers; diagrammatic, minimal, educational, pastel, interface-inspired - Length and format: 2 min, 874:720 (874x720 source) - Playback: a muted visual reference. The original soundtrack is available only at the source. - Shared by @joaoli13 in X, Sep 23, 2026: https://x.com/joaoli13/status/2102805376466891137 - Prompt fidelity: Verbatim. This is the complete prompt exactly as the creator published it. ## Before you start Guidance from Reference, separate from the creator's instructions. 1. Read the reference and inspect the project, installed libraries, and available agent skills before writing code. Reuse suitable tools already present. 2. Prefer the creator-reported libraries and skills when they fit the task. Explain what useful additions help with, and install missing project dependencies or relevant skills from official sources when the environment and the user's existing permissions allow it. Respect the project's package manager and setup rules; do not ask again for routine setup already authorized. 3. Distinguish tools used by the creator from your own recommendations. A listed tool is not automatically required, and an empty tools or requirements list means the setup is unknown, not that nothing is needed. 4. Ask concise questions when missing inputs or choices materially affect the result: paid services, credentials, the user's assets, or a different production approach. Use existing answers and permissions; continue independent work while waiting. 5. If a useful library, skill, or service cannot be installed or used, explain an available alternative and its expected effect on quality, features, or effort. Get the user's choice before a material compromise; otherwise state the fallback and continue. Never silently simplify the result or claim to have used a tool you did not use. Requirements reported by the source: Install the creator-stated skill or plugin: JavaScript.. ## Recreating this video These are the creator's published instructions, unedited. ## Prompts ### 1. Context-window explainer prompt ```text Faça uma animação javascript de como um LLM vai compondo na janela de contexto as perguntas do usuário e as respostas geradas. É importante diferenciar a janela de contexto de entrada da de saída, que normalmente possuem tamanhos diferentes, e deve-se mostrar também que a resposta que foi gerada na janela de contexto de saída é transferida e concatenada ào contexto da janela de entrada. Deve ser claro e didático e pode ter labels em português ``` ### 2. MP4 export request ```text gere a animação no formato MP4 para download ``` ## What it looks like Written by Reference from the video frames, not by the creator. A Spanish-language educational interface explaining an LLM context window. - Color: Pale blue-gray background with white panels, blue outlines, and teal and orange highlights. - Type: Small, clean sans-serif text with bold headings and compact labels. - Framing: A dashboard-like arrangement of explanatory text, a token grid, an LLM panel, a database panel, and a side column of notes. - Structure: The main content is organized into bordered sections: an introduction, a context-window grid, an LLM diagram, a database area, and supporting callouts. Storyboard, nine frames spread across the video: https://meetreference.com/media/video-llm-context-windows-and-token-grids-b2a86d/storyboard.webp - Frame 1 at 5.6s: https://meetreference.com/media/video-llm-context-windows-and-token-grids-b2a86d/f1.webp - Frame 2 at 16.9s: https://meetreference.com/media/video-llm-context-windows-and-token-grids-b2a86d/f2.webp - Frame 3 at 28.1s: https://meetreference.com/media/video-llm-context-windows-and-token-grids-b2a86d/f3.webp - Frame 4 at 39.3s: https://meetreference.com/media/video-llm-context-windows-and-token-grids-b2a86d/f4.webp - Frame 5 at 50.6s: https://meetreference.com/media/video-llm-context-windows-and-token-grids-b2a86d/f5.webp - Frame 6 at 61.8s: https://meetreference.com/media/video-llm-context-windows-and-token-grids-b2a86d/f6.webp - Frame 7 at 73.0s: https://meetreference.com/media/video-llm-context-windows-and-token-grids-b2a86d/f7.webp - Frame 8 at 84.3s: https://meetreference.com/media/video-llm-context-windows-and-token-grids-b2a86d/f8.webp - Frame 9 at 95.5s: https://meetreference.com/media/video-llm-context-windows-and-token-grids-b2a86d/f9.webp ## Links - Watch: https://x.com/joaoli13/status/2102805376466891137 - Original post: https://x.com/joaoli13/status/2102805376466891137 - Discovery collection: https://data.openchamber.dev/opus-5-5-videos/v/joaoli13-2102805376466891137