Hackernews posts about Benchmarks

1. 61
   

   newBtrfs/ZFS/bcachefs under workloads classic benchmarks skipbartosz.fenski.pl

   ▲ 179farlight15d💬 190savesaved
2. 62
   

   newRTK reports token savings, but our cost benchmarks disagreequesma.com

   ▲ 170michalwarda23d💬 84savesaved
3. 63
   

   newShow HN: JevBench, a reproducible benchmark for typed decision modelsbenchmarkheaven.com

   ▲ 154florianstandhar12d💬 39savesaved
4. 64
   

   newRecreating Minecraft Is Not a Benchmarkkuber.studio

   ▲ 73kuberwastaken28d💬 74savesaved
5. 65
   

   newBad benchmarks and evals: Senior SWE-Bench, napkin math, and winter tiresdanluu.com

   ▲ 56luu23d💬 46savesaved
6. 66
   

   newBenchmark: CadQuery vs. OpenSCAD for agentic CAD workmodelrift.com

   ▲ 40jetter22d💬 65savesaved
7. 67
   

   newUncensored and Offensive Security AI Models Benchmarkgithub.com

   ▲ 38soltanov5d💬 12savesaved
8. 68
   

   new1Password's AI patching benchmark is misleadingblog.trailofbits.com

   ▲ 37darkamaul19d💬 9savesaved
9. 69
   

   newAgents on Rails: Best model solves 35% of feature benchmark runsrubyonrails.org

   ▲ 20chalmovsky23d💬 4savesaved
10. 70
    

    newDoGBench: The first user-facing docs generation benchmark. No model scores >50%dogbench.ai

    ▲ 18prithvi22063d💬 3savesaved
11. 71
    

    newAuto-autoresearch: self-improving agents on Karpathy's NanoChat benchmarkrekursiv.ai

    ▲ 13hyperparticle19d💬 5savesaved
12. 72
    

    newThe case to BYOB: build your own (coding) benchmarksbyobench.ai

    ▲ 11oaa3624d💬 2savesaved
13. 73
    

    newThe Two MMLU Scores: What a Benchmark Name Does Not Fixzatona.dev

    ▲ 8gmays21d💬 3savesaved
14. 74
    

    newInsurance Agent Benchmark: 166 real-world cases for evaluating insurance AIwww.askcooper.ai

    ▲ 7jsc3917ddiscusssavesaved
15. 75
    

    newKernelBench: Can LLMs Write GPU Kernels? – Benchmark and Toolkit, Torch –> CUDAgithub.com

    ▲ 6peter_d_sherman3ddiscusssavesaved
16. 76
    

    newWhy Bend 2's typechecker benchmarks are misleading (and its practical flaws)gist.github.com

    ▲ 6Nezk16ddiscusssavesaved
17. 77
    

    newShow HN: Benchmark: AI doesn't find bugs unless you tell it what's wrongswesweep.com

    ▲ 5lieret3d💬 2savesaved
18. 78
    

    newiOS and macOS CI benchmark: Avrea is 3-23x faster than GitHub Actionsavrea.com

    ▲ 5leolannenmaki3d💬 2savesaved
19. 79
    

    newShow HN: Livenerf – a benchmark for whether Opus 5.5 gets nerfedgithub.com

    ▲ 5ninjahawk112d💬 2savesaved
20. 80
    

    newShow HN: Compute:Arena – Community submitted local AI benchmarkscomputearena.ai

    ▲ 5prabod17d💬 2savesaved
21. 81
    

    newAsk HN: Model sycophancy benchmarks?

    ▲ 5gradus_ad28d💬 2savesaved
22. 82
    

    newShow HN: A benchmark comparison of C++ vs. Node.js performancevikky2810.github.io

    ▲ 5vikky281010d💬 1savesaved
23. 83
    

    newBecoming a Benchmarkterrytao.wordpress.com

    ▲ 5bluepeter17ddiscusssavesaved
24. 84
    

    newBenchmarks are more broken than we could have imaginedwww.horizonanalyticslabs.com

    ▲ 4chillacy12d💬 1savesaved
25. 85
    

    newThe Cure for Cancer Is Not a Pure Math Benchmarkscholars-stage.org

    ▲ 4zdw15d💬 1savesaved
26. 86
    

    newMoorcheh Edge on Arduino UNO Q vs. Ventuno Q: 1536-D Vector Benchmarkmoorcheh.ai

    ▲ 4supportm2ddiscusssavesaved
27. 87
    

    newgo/codegolf benchmark: frontier models vs. Google's 2015 code golfersbjin.github.io

    ▲ 4bjin6ddiscusssavesaved
28. 88
    

    newM5 Ultra first benchmarks – Geekbenchbrowser.geekbench.com

    ▲ 4happyopossum19ddiscusssavesaved
29. 89
    

    newThe Benchmark Creator Is Almost Always on the Podiumpymupdf.io

    ▲ 4akaybk19ddiscusssavesaved
30. 90
    

    newLLM Benchmark for de-identification and synthesishuggingface.co

    ▲ 3akamor11d💬 2savesaved

· edit