The best AI benchmarks, according to hackers
10 AI benchmarks were recommended on Hacker News between November 2025 and November 2026. We rank them by how many separate people recommend each one, and 1 earns the Hacker Approved badge.
Top pick
The most recommended AI benchmark with the badge. Each figure is one person. RecommendsCriticizes One comment
Side-by-side comparison of OCR models
Free place to upload a document and compare OCR models side by side. Critics question its ELO rankings, which lack error bars.
All 10 AI benchmarks, ranked
Select a column heading to sort. Approval share is recommenders out of everyone who gave an opinion; the tick on each bar marks the 2 to 1 line the badge needs. Months counts the months, out of 13, in which someone recommended it.
| Rank | |||||||
|---|---|---|---|---|---|---|---|
| 1 | Artificial Analysis | 62 | 46 | 57% | 11 | — | — |
| 2 | LMArena | 39 | 40 | 49% | 11 | — | — |
| 3 | ARC-AGI | 22 | 24 | 48% | 8 | — | — |
| 4 | OCR Arena | 11 | 4 | 73% | 4 | — | Approved |
| Tied 5 | Bullshit Benchmark | 9 | 0 | 100% | 6 | — | — |
| Tied 5 | GenAI Showdown | 9 | 3 | 75% | 5 | — | — |
| Tied 5 | CivBench | 9 | 1 | 90% | 1 | — | — |
| Tied 8 | Artificial Analysis Intelligence Index | 8 | 49 | 14% | 3 | — | — |
| Tied 8 | AI World Clocks | 8 | 5 | 62% | 1 | — | — |
| 10 | Brood War Bench | 7 | 0 | 100% | 1 | — | — |
How we rank
Rank is the number of separate people who recommend a product in a Hacker News comment over the past year. Someone who praises the same AI benchmark in five threads counts once.
The badge goes to AI benchmarks that at least 10 people recommend, with recommenders outnumbering critics at least 2 to 1, recommended in at least 4 of the 13 months.
Rankings are never for sale. We may earn a commission when you buy through our links. It never changes the order.