Cardio Chase

Company: Cardio Chase

Self-reported basis: The description is entirely self-reported and unverified, based on an author-submitted write-up for the OpenAI 2026 hackathon. No external corroboration or historical data exists.

What it appears to be: A privacy-first macOS fitness game that uses body tracking via camera input to control gameplay, turning real physical activity into in-game movement.

What changed: The project evolved from a camera feasibility spike into a playable v1.0 release with deterministic movement recognition, local-first architecture, and support for both front and side camera inputs.

Most important open question: Is the product’s motion recognition robust enough to be reliable across diverse physical conditions, or does it fail in ways that prevent consistent gameplay?

The description states that Cardio Chase is a native macOS fitness game built using Swift and SwiftUI, with SpriteKit for 2.5D rendering. It uses two camera inputs — one from the Mac’s front camera and another via iPhone Continuity Camera placed at the side — to track player movement. The app tracks upright play, jumps, and push-ups through local body pose detection, using Apple Vision to extract body joints from both streams.

Movement recognition is implemented using state machines that interpret lane position, cadence, deliberate jumps, and full push-up sequences. The game does not upload or store data; all processing and persistence happen locally on the Mac.

• Evidenced: It is a macOS app built with Swift, SwiftUI, SpriteKit, AVFoundation, and Apple Vision.
• Inferred: The game is designed to be a fitness-enhanced endless runner with physical activity as core gameplay resource.
• Not evidenced: No revenue, customers, or usage data.

The author states that Cardio Chase aims to bridge the gap between motion games (which often use gestures) and fitness apps (which struggle to make cardio enjoyable). It positions itself as a privacy-first fitness game, where physical effort directly controls gameplay, and no data is uploaded or stored.

• Evidenced: The app is described as privacy-first, with no server-side processing.
• Inferred: The positioning is that of a hybrid between gamification and fitness, using real movement to drive game progression.
• Not evidenced: No claims about market traction, user adoption, or competitive differentiation beyond its own description.

The author does not explicitly define a target customer or ideal customer profile (ICP). The app is described as a fitness game for macOS users, with support for both front and side camera inputs. It appears to be designed for individuals who are interested in fitness, physical activity, and gamified experiences.

• Evidenced: The app targets macOS users who engage in cardio exercises like jogging, jumping, and push-ups.
• Inferred: Likely a niche audience of fitness enthusiasts or developers interested in motion-based apps.
• Not evidenced: No stated customer segments, personas, or user demographics.

The description does not mention any pricing model, monetization strategy, or business model. It is described as a v1.0 release and a hackathon submission, with no indication of paid features, subscriptions, or in-app purchases.

• Evidenced: No pricing or monetization details.
• Inferred: Likely free-to-play or one-time purchase, but this is not stated.
• Not evidenced: No revenue model, pricing tiers, or commercial strategy.

The app is built as a native macOS 15 SwiftPM project using SwiftUI and SpriteKit. It uses AVFoundation to run two independent camera sessions, and Apple Vision for local body joint detection. The system normalizes poses from both cameras and combines semantic results within a timestamp window.

• Evidenced: Uses Swift, SwiftUI, SpriteKit, AVFoundation, Apple Vision.
• Inferred: The architecture is designed for local-first processing with deterministic movement recognition.
• Not evidenced: No details on scalability, performance metrics, or deployment strategy beyond the v1.0 release.

The project is described as a v1.0 release and was submitted to the OpenAI 2026 hackathon. It includes features like a tutorial, deterministic progression, original assets, and support for external or built-in cameras.

• Evidenced: v1.0 release, built-in tutorial, original assets.
• Inferred: The app is mature enough for a hackathon submission but lacks real-world usage data.
• Not evidenced: No user base, retention metrics, or adoption data.

The author does not reference any competitors or market context. The description implies that Cardio Chase fills a gap between motion games and fitness apps, but no comparison to existing products is made.

• Evidenced: No mention of competitors.
• Inferred: Likely in a niche space with limited direct competition.
• Not evidenced: No competitive analysis or positioning relative to other apps.

• Motion recognition robustness: The author notes challenges like camera occlusion, disconnected cameras, and stale push-up states. If these issues are not fully resolved, gameplay may be inconsistent or frustrating.
• Limited testing: The app is described as tested by the single developer in a limited environment; broader real-world validation is missing.
• Privacy model: While privacy is emphasized, the lack of server-side processing raises questions about how user data is handled and whether it can scale to more complex features.
• No monetization strategy: The app is not described as monetized, which may limit long-term viability.

1. How does the app handle camera failures or occlusions during gameplay?
2. What are the limitations of the current motion recognition in real-world conditions?
3. Has the app been tested with a diverse group of users under various lighting and physical conditions?
4. Are there plans to expand beyond macOS or support other platforms?
5. What is the long-term vision for monetization or user engagement beyond the initial release?