Mirrorcode
The Frontier
Best score over time · one chart, every benchmark
Full rankings
8 models tested · sorted by score
| # | Model | Score |
|---|---|---|
| 1 | 77.4 | |
| 2 | 73.3 | |
| 3 | 63.9 | |
| 4 | 46.7 | |
| 5 | 31.1 | |
| 6 | 20.0 | |
| 7 | 15.6 | |
| 8 | 8.9 |
Score distribution
Where models cluster
Correlated benchmarks
Pearson r · original research
Benchmarks that track with Mirrorcode
Pearson correlation across models scored on both benchmarks. Closer to 1 = strongly predictive.
Frequently asked
About Mirrorcode
What does Mirrorcode measure?
Mirrorcode is a knowledge benchmark in the BenchGecko catalog. 8 AI models have been tested on it. Scores range from 8.9 to 77.4 out of 100.
Which model leads on Mirrorcode?
Claude Opus 5.5 from Anthropic leads Mirrorcode with a score of 77.4. The median score across 8 tested models is 38.9.
Is Mirrorcode saturated?
No · the top score is 77.4 out of 100 (77%). There is still meaningful room for improvement on Mirrorcode.
Does Mirrorcode predict performance on other benchmarks?
Yes · Mirrorcode scores correlate 0.92 with APEX-Agents across 8 shared models. Models that do well on Mirrorcode tend to do well on APEX-Agents.
How often is Mirrorcode data refreshed?
BenchGecko pulls updates daily. New model scores on Mirrorcode appear as soon as they are published by Epoch AI or the model provider.
- Category
- Knowledge
- Max score
- 100
- Models
- 8
- Updated
- 2026-09-22
Top on Mirrorcode
Claude Opus 5.5 · 77.4Claude Fable 5.1 · 73.3Claude Fable 5 · 63.9GPT-6 Astra · 46.7Claude Opus 4.7 · 31.1More knowledge benchmarks
Same category · related evaluations