Subquadratic · Launch Video Breakdown: Hook, Pacing & Motion Design
The first model built for long‑context tasks
Scene-by-scene timeline & spoken transcript
The Hook
Breakthrough Architecture Reveal
“Today we're announcing a major architectural breakthrough. The world's first fully subquadratic LLM. SubQ has a 12 million token context window, and frontier-level intelligence.”
- On screen
- Alex Whedon Co-Founder & CTO World's first fully subquadratic LLM 12,000,000 tokens
- Camera
- Static mid-shot framed by dynamic colored pixel mosaic blocks.
- Motion
- Pixel-block border reveal with kinetic typographic wipes.
Product Reveal
Benchmark Dominance
“It outperforms Opus 4.7 on long context at less than 5% of the cost.”
- On screen
- SWE-bench verified Agentic Coding MRCR v2 Long-context retrieval at 1M tokens
- Camera
- Static graphic plane.
- Motion
- Animated bar charts with highlighted benchmark deltas.
Feature Teaser
Speed and Scaling Milestones
“At 1 million tokens, SubQ processes tokens 52 times faster than Flash Attention. It dramatically changes the cost of training and inference, enabling us to scale context windows to 12 million tokens and beyond, pushing the boundaries of intelligence.”
- On screen
- 52X Faster 52x faster | Flash Attention
- Camera
- Static medium close-up interspersed with motion graphic card.
- Motion
- Pixelated multiplier counter and clean typographic wipe.
Feature Teaser
Agent Workflows at Scale
“This means your agents can now complete weeks of work at a time without degrading. Reasoning across entire codebases, merging hundreds of PRs at once, and finding patterns in tens of thousands of documents. All in one shot, without losing accuracy, speed, or context.”
- On screen
- Reasoning codebases merging hundreds at once finding patterns of thousands accuracy, speed
- Camera
- Rapid 2D graphic cuts.
- Motion
- Calendar checklists, terminal code scrolling, merging PR cards, grid arrays, and reticle zoom.
Product Reveal
Sparse-Attention Core
“SubQ is not just another model, it represents a major algorithmic breakthrough. It is the first model built on a fully sub-quadratic, sparse-attention architecture.”
- On screen
- (No on-screen text)
- Camera
- Static medium shot with bookshelf backdrop.
- Motion
- Live founder presentation.
Problem Agitation
Standard vs Sparse Attention
“LLMs today waste compute by processing every possible relationship between words. But only a small fraction of these matter. SubQ finds and focuses only on those relationships, meaning compute is used where it matters most. At 12 million tokens, this reduces compute almost 1,000 times, changing the way that LLMs scale.”
- On screen
- Standard Attention SubQ Sparse Attention 1000x
- Camera
- Diagram overlay tracking speaker.
- Motion
- Circular node network animating from dense cross-attention to sparse active pathways.
Call to Action
Access Announcement & Outro
“Today, SubQ is available for early access, along with SubQ Code, a coding agent built on the next generation of AI technology. This marks a new era for builders. An era powered by subquadratic intelligence. Visit Subq.ai to get early access.”
- On screen
- Visit Subq.ai to get early access.
- Camera
- Static medium shot resolving to black title card.
- Motion
- Color bar sweep transitioning to 2D pixel mosaic logo lockup.
Motion design primitives
Synthesizable in Remotion · key components of the subquadratic-launch-explainer template.
AttentionGraph
Circular node cluster transitioning from full radial edges to sparse selective links.
+ Code construction recipe
Generate node positions along an SVG circle. Render connecting lines whose opacity interpolates based on Remotion frame state.
PixelFrameBorder
Multi-colored square blocks populating the outer screen perimeter.
+ Code construction recipe
Build a flex grid border populated with animated color cells using spring-driven opacity and scale.
BenchmarkBarSequence
Side-by-side vertical bar graphs rising to distinct target heights with top label tags.
+ Code construction recipe
Animate SVG rect height using interpolate() with spring damping, rendering percentage tags above the final heights.
Style DNA & aesthetic system
Visual aesthetic
Typography
Motion
Special effects
Pacing & rhythm
Conceptual story rhythm
- 1
Breakthrough Metric
Purpose. Hook technical audiences with state-of-the-art context window specs.
Execution. Presenter speaking framed by colorful pixel-grid edges and large context numbers.
- 2
Empirical Proof
Purpose. Validate model strength using standard industry benchmarks.
Execution. Dark-mode bar graphs showing SWE-bench and long-context retrieval scores.
- 3
Architectural Mechanism
Purpose. Explain how sparse attention enables 1000x compute reduction.
Execution. Interactive node graph shifting from wasteful dense connections to pruned sparse paths.
- 4
Call to Action
Purpose. Direct users to register for early access.
Execution. Clean serif text URL over presenter resolving to the pixel logo on black.
Remix in Animatiq Studio
Master remix prompt · agent-ready
Sample remix
Objective
Announce a novel vector database engine featuring 100x query speeds.How the recipe adapts
Replace the attention graph with a high-dimensional spatial indexing visualization and calibrate the bar charts to vector throughput benchmarks.Remix blueprint · subquadratic-launch-explainer
Scene durations (7 scenes, 85s)
Key Remotion components
Related AI Agents Startup Launches
Explore all 1064 AI Agents launches →
To ensure that artificial general intelligence benefits all of humanity
Introducing Claude Fable 5: a Mythos-class model that we’ve made safe for general use. Its capabilities exceed those of any model we’ve ever made generally available. https://t.co/2AvmEjHIX8
To understand the true nature of the universe.