GBA Eval - Gameplay Replays: leaderboard
GBA Eval section score for Game Boy Advance emulators that models write in Rust/WebAssembly: recorded gameplay inputs are replayed and frames compared with the Mesen2 reference by SSIM, 0-1, higher is better. Carries 60% of the overall score; ungraded emulators are omitted.
Metric: Frame-by-frame SSIM on replayed gameplay inputs (0-1). Source: gbaeval.com. Saturation forecast: Around February 2027. 21 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 5 | 0.93 | #16 |
| 2 | Claude Fable 5 | 0.92 | #20 |
| 3 | Claude Opus 4.8 | 0.77 | #39 |
| 4 | Grok 4.5 | 0.77 | #54 |
| 5 | Claude Sonnet 5 | 0.71 | #65 |
| 6 | GPT-5.5 | 0.61 | #33 |
| 7 | Claude Sonnet 4.6 | 0.59 | #94 |
| 8 | Claude Opus 4.6 | 0.59 | #67 |
| 9 | GPT-5.6 Sol | 0.56 | #22 |
| 10 | Kimi K3 | 0.55 | #35 |
| 11 | GPT-5.4 | 0.53 | #83 |
| 12 | Claude Opus 4.7 | 0.52 | #51 |
| 13 | Muse Spark 1.1 | 0.1 | #55 |
| 14 | Gemini 3.5 Flash | 0.05 | #66 |
| 15 | Grok Build 0.1 | 0.02 | #137 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=gba-eval-gameplay-replays · How It Works · Data refreshed daily, snapshot 2026-11-05.