Hackernews posts about Claude Sonnet 4

  1. Claude Sonnet 4.6 says it is 我是 DeepSeek when asked in Chinese (xcancel.com)
    10 points by pera 21 days ago | discuss
  2. Knuth Test using Claude Sonnet 4.6 problem 1.1.3
    3 points by daly 15 days ago | discuss
  3. We ran 21 MCP database tasks on Claude Sonnet 4.6 (insforge.dev)
    2 points by Arindam1729 12 days ago | discuss
  4. Knuth Test Using Claude Sonnet 4.6 Problem 1.1.2
    1 points by daly 15 days ago | discuss
  5. Knuth Test Using Claude Sonnet 4.6 problem 1.1.1
    1 points by daly 15 days ago | discuss
  6. Knuth Tests using Claude Sonnet 4.6 problem 1.1.4
    1 points by daly 15 days ago | discuss
  7. “Car Wash” test with 53 models (opper.ai)
    371 points by felix089 25 days ago | 447 comments
  8. Show HN: Keep large tool output out of LLM context: 3x accuracy 95% fewer tokens (github.com)
    10 points by loumaciel 16 days ago | 1 comments
  9. If you drive clock wise along the beach on an island
    7 points by Cookingboy 22 days ago | 5 comments
  10. Show HN: Cloudwright – validate, cost, and export cloud architectures from text (github.com)
    6 points by xmpuspus 17 days ago | discuss
  11. Show HN: Pencil Puzzle Bench – LLM Benchmark for Multi-Step Verifiable Reasoning (ppbench.com)
    5 points by bluecoconut 17 days ago | discuss
  12. Show HN: I made a static site for exploring names (namex.lyall.co)
    4 points by lyall 29 days ago | 8 comments
  13. Show HN: 58% cost by replacing file reads with a dependency graph on AI Coding
    4 points by nicola_alessi 18 days ago | 2 comments
  14. Show HN: Ablo - AI slides without the generic look or layout restrictions (www.ablo.finance)
    3 points by lukasanderss 18 days ago | 2 comments
  15. Show HN: FC-Eval – CLI to Benchmark Local or Cloud LLMs on Function Calling (github.com)
    3 points by gauravvij137 4 days ago | discuss
  16. Show HN: I added barcode scanning to my AI pricing app – 92% cheaper per scan (underpricedai.com)
    2 points by fkratzer 7 days ago | 1 comments
  17. Show HN: Lumen – vision-first browser agent (state of the art, open source) (github.com)
    2 points by fearlessboi 10 days ago | 1 comments
  18. The Car Wash Problem: A variable isolation study on prompt architecture
    2 points by midmost44 29 days ago | 1 comments
  19. Show HN: Caliper – Auto Instrumented LLM Observability with Custom Metadata (gitlab.com)
    2 points by OliverGuy 13 days ago | discuss
  20. Show HN: Focused input cuts LLM output tokens by 63% bench on CC with FastAPI
    2 points by nicola_alessi 17 days ago | discuss
  21. More tokens, less cost: why optimizing for token count is wrong
    1 points by nicola_alessi 10 days ago | 9 comments
  22. Show HN: I built an AI-powered technical interview prep tool (crackr.dev)
    1 points by wa5ina 11 days ago | 1 comments
  23. Show HN: We attacked 8 AI agents with adversarial probes – 5 failed (antarraksha.ai)
    1 points by akash1039ai 17 days ago | discuss
  24. Claude Opus 4.6 vs. Sonnet 4.6 Coding Comparison (www.tensorlake.ai)
    1 points by shricodevvvv 16 days ago | discuss
  25. 1M context window now generally available for Claude Opus and Sonnet 4.6 (twitter.com)
    6 points by tosh 7 days ago | discuss
  26. Show HN: Altimate Code – Open-Source Agentic Data Engineering Harness (github.com)
    20 points by aaur0 1 day ago | 2 comments
  27. Show HN: Yardstiq – Compare LLM outputs side-by-side in your terminal (www.yardstiq.sh)
    2 points by stanleycyang 17 days ago | discuss
  28. Show HN: Porchsongs.ai; Rewrite chordcharts/lyrics with AI to make them personal (porchsongs.ai)
    1 points by river_otter 16 days ago | discuss
  29. How long till every major provider sets their RSI loops in motion?
    3 points by foxindustrial 9 days ago | discuss
  30. Launch HN: Canary (YC W26) – AI QA that understands your code
    58 points by Visweshyc 1 day ago | 25 comments
More