Hackernews posts about Benchmarks

1. Meta got caught gaming AI benchmarks (www.theverge.com)
   347 points by pseudolus 14 days ago | 161 comments
2. Show HN: LocalScore – Local LLM Benchmark (www.localscore.ai)
   124 points by sipjca 19 days ago | 25 comments
3. LLM Benchmark for 'Longform Creative Writing' (eqbench.com)
   96 points by vitorgrs 12 days ago | 88 comments
4. New open-source benchmark for real-time analytics applications (github.com)
   16 points by thenoahhein 26 days ago | 4 comments
5. Vending-Bench: A Benchmark for Long-Term Coherence of Autonomous Agents (arxiv.org)
   14 points by distalx 2 days ago | discuss
6. Medical Benchmarks and the Myth of the Universal Patient (www.newyorker.com)
   12 points by pseudolus 26 days ago | 3 comments
7. DeepSeek-V3-0324 Crushes GPT-4.5 in Math and Code Benchmarks at 1/277 the Cost (api-docs.deepseek.com)
   10 points by kong75 27 days ago | 2 comments
8. LocalScore: A Local LLM Benchmark (www.localscore.ai)
   9 points by jborichevskiy 18 days ago | 2 comments
9. DeepSeek V3 0324 outpaces GPT 4.5 and Claude 3.7 in coding, other benchmarks (huggingface.co)
   7 points by xianshou 26 days ago | discuss
10. NPB-Rust: NAS Parallel Benchmarks in Rust (arxiv.org)
    6 points by aw1621107 7 days ago | 1 comments
11. Show HN: CVE-Bench, the first LLM benchmark using real-world web vulnerabilities (github.com)
    6 points by jbenn 21 days ago | 1 comments
12. BodoBench95: A Modern Benchmark for the Timeless Power of the Intel Pentium Pro (www.bodo.ai)
    5 points by marquisdepolis 21 days ago | 1 comments
13. DuckDB's CSV Reader and the Pollock Robustness Benchmark: Into the CSV Abyss (duckdb.org)
    5 points by CipherDrift 4 days ago | discuss
14. Llama 4 underperforms: a benchmark against coding-centric models (rootly.com)
    5 points by sylvainkalache 8 days ago | discuss
15. BrowseComp: A Benchmark for Browsing Agents (openai.com)
    5 points by meetpateltech 12 days ago | discuss
16. Meta exec denies the company artificially boosted Llama 4's benchmark scores (techcrunch.com)
    5 points by pera 14 days ago | discuss
17. Meta's benchmarks for its new AI models are a bit misleading (techcrunch.com)
    5 points by sherlockxu 15 days ago | discuss
18. A16Z- and Benchmark-backed 11x has been claiming customers it doesn't have (techcrunch.com)
    4 points by nradov 27 days ago | 2 comments
19. Tpchgen-rs: TPC-H benchmark data generation in pure Rust (github.com)
    4 points by todsacerdoti 10 days ago | 1 comments
20. The Disconnect Between AI Benchmarks and Math Research (sugaku.net)
    4 points by rfurmani 28 days ago | discuss
21. O4-mini-high passed my personal benchmark hallucinations, so here it is (twitter.com)
    3 points by nailer 4 days ago | 1 comments
22. LiveBench: A Challenging, Contamination-Free LLM Benchmark (livebench.ai)
    3 points by foolswisdom 13 days ago | discuss
23. Meta cheats on Llama 4 benchmark (www.heise.de)
    3 points by Topfi 13 days ago | discuss
24. Show HN: The SQLite Drivers 25.04 Benchmarks Game (Golang) (pkg.go.dev)
    3 points by 0xjnml 18 days ago | discuss
25. Show HN: Docsumo's OCR Benchmark Report – Surpassing Mistral and Landing AI (www.docsumo.com)
    3 points by snehanairdoc 19 days ago | discuss
26. RTX 5090 Mobile: First LLM Benchmarks Are In (www.hardware-corner.net)
    3 points by mdp2021 24 days ago | discuss
27. Benchmark Comparison of Rust Logging Libraries (github.com)
    3 points by jacksonshi 29 days ago | discuss
28. OpenAI o3 AI model scores lower on benchmark than the company initially implied (techcrunch.com)
    2 points by rntn about 20 hours ago | discuss
29. InternVL3: Open-Source Model Outperforms GPT-4o in Multimodal Benchmarks (www.implicator.ai)
    2 points by marcusschuler 7 days ago | discuss
30. Show HN: Benchstats – Statistical Testing for Benchmark Results Comparison (github.com)
    2 points by Arech 10 days ago | discuss