Eye Tracking Feasibility: Godot vs Unity vs Unreal

emerging Updated: 2026-05-31
godot unity unreal eye-tracking comparison licensing feasibility device-access research

Eye Tracking Feasibility: Godot vs Unity vs Unreal

Status: emerging
Last updated: 2026-05-31
Sources: Engine Comparison Eye Tracking Feasibility
Tags: [godot, unity, unreal, eye-tracking, comparison, licensing, feasibility, device-access, research]

Summary

A May 2026 feasibility comparison ranks the three engines for eye-tracking research across cost, device data access, and result reliability. On cost, all three are effectively free for non-commercial research, with Godot carrying zero licensing risk at any scale (Anonymous, 2026). On data access depth and on research reliability, the ranking is Unity > Unreal > Godot, driven mainly by Unity's Varjo and SRanipal coverage and its established use in published VR studies (Anonymous, 2026). The headline recommendation for academic gaze research is Unity paired with Varjo XR-4 hardware. The comparison comes from a single unverified digest, so its rankings are treated as emerging.

Body

Context

This article draws on a single May 2026 feasibility comparison of Godot, Unity, and Unreal for eye-tracking research (Anonymous, 2026), an unverified web-search digest. The source ranks the three engines along three axes — monetary cost, depth of device data access, and reliability of results — and ends with use-case recommendations. Within this knowledge base it is the synthesis article of the eye-tracking-in-game-engines strand: it sits above the three per-engine surveys (Godot Eye Tracking, Unity Eye Tracking, Unreal Eye Tracking) and draws the shared baseline from Openxr Eye Gaze Interaction and the SRanipal risk factors from Sranipal Sdk into one cross-engine ranking. Because the comparison comes from one unverified digest, its rankings are treated as emerging throughout.

Key Points

On monetary feasibility the three engines converge for research but diverge at commercial scale. Godot is MIT-licensed with no royalties, revenue thresholds, per-seat fees, or runtime fees, so it is zero-cost at every scale; Unity is free on Personal up to $200K revenue/funding (Pro at $2,310/seat/year above that, runtime fee cancelled September 2024, free Education Grant for institutions); and Unreal is free for non-commercial research and for commercial revenue below $1M, with a 5% royalty only above $1M plus Epic MegaGrants for XR research. The digest ranks monetary feasibility Godot ≥ Unreal > Unity, while noting that for pure research with no commercial revenue all three are effectively free and Godot alone carries no future licensing risk (§ Dimension 1: Monetary Feasibility). The per-engine licensing detail is in Unity Eye Tracking and Unreal Eye Tracking.

On depth of device access the ranking inverts to Unity > Unreal > Godot. Unity reaches the broadest, most accessible native data across hardware families, especially the Varjo plus SRanipal combination, with Varjo XR-4 plus the Varjo Unity plugin yielding calibrated millimetre pupil and iris diameter, 200 Hz, and nanosecond timestamps without licensing beyond hardware. Unreal is close behind, with per-eye gaze through several plugins and normalized pupil via VIVE OpenXR on UE5, though Varjo's calibrated mm metrics are reachable mainly through C++. Godot is the most limited: combined gaze direction via OpenXR everywhere, per-eye only through the Meta Toolkit on Quest Pro/3, and no pupil diameter, no blink via OpenXR, and no sample rate above 120 Hz through any current plugin (§ Dimension 2: Depth of Device Access (Without Third-Party Limitations)). These per-engine data sets are detailed in Godot Eye Tracking, Unity Eye Tracking, and Unreal Eye Tracking.

On reliability of results the digest again ranks Unity > Unreal > Godot, on validation, timing, and ecosystem maturity. Unity has multiple published VR-with-eye-tracking studies (commonly Varjo XR-3/4 plus Unity and Quest Pro plus Unity), more predictable frame timing, Varjo nanosecond timestamps that can anchor stimuli to gaze samples, and published Varjo data-quality specs (~0.4° accuracy, ~0.07° precision, 200 Hz) with independent validation. Unreal has fewer published studies, more variable UE5 frame times, and a Blueprint abstraction layer that can add latency versus direct C++ access. Godot has no published studies found as an eye-tracking stimulus platform as of May 2026, no research-grade frame timing, and no built-in data logging or event-marker system (§ Dimension 3: Reliability of Results). The "no studies found" and timing claims are single-source and treated as emerging.

The recommendations follow from these rankings: Unity plus Varjo XR-4 for academic gaze/psychophysiology research (best data quality, published precedent, calibrated mm pupil, 200 Hz); Godot for cost-sensitive prototypes (zero cost, basic gaze, accepting data limits); Unreal for no-budget non-commercial VR research (fully free, good per-eye plus openness via VIVE OpenXR); and either Unity or Unreal for XR games depending on team expertise (§ Overall Recommendation by Use Case). Cross-engine risk factors recur: SRanipal's maintenance mode, lack of UE5 support, and normalized-only pupil (see Sranipal Sdk); Tobii Ocumen's limited hardware and commercial license; the insufficiency of OpenXR-only gaze for pupillometry or high-frequency work (see Openxr Eye Gaze Interaction); and the untested-in-research status of Godot custom bridges (§ Key Risk Factors (All Engines)).

Conclusion

Set side by side, the three engines occupy distinct positions. Unity is the most research-mature: deepest device access, the only calibrated-mm-and-200-Hz path (via Varjo), and the only one with published eye-tracking study precedent, at the cost of the most complex licensing at commercial scale. Unreal sits close behind on data and is the most permissive for no-budget work, but trails on published precedent and timing predictability and routes its richest Varjo metrics through C++. Godot is the cheapest and licence-clean at any scale, but the most limited in data and the only one with no published research use found and no built-in logging or event markers. The remaining gaps are the absence of independent validation for Godot research use and the dependence of all three rankings on a single unverified digest, so they should be read as a structured hypothesis rather than a validated finding.

References

Anonymous (2026) Godot vs Unity vs Unreal — Eye Tracking Feasibility Comparison (Web Search Digest). Research digest, virtual-environments-kb RAW (Accessed: 31 May 2026). engineComparisonET2026

Open Questions

  • Are the comparative rankings reproducible against primary documentation, given the single unverified source?
  • Has any Godot eye-tracking research been published since May 2026 that would change the reliability ranking?
  • Do the cited Varjo data-quality specs (~0.4° accuracy, ~0.07° precision) trace to an independent validation study that could be ingested directly?