https://hardnumbers.dev2026-09-03T22:26:13+00:00daily1.0 https://hardnumbers.dev/articles2026-09-03T22:26:13+00:00daily0.9 https://hardnumbers.dev/about2026-09-03T22:26:13+00:00monthly0.5 https://hardnumbers.dev/pillar/agents2026-09-03T22:26:13+00:00weekly0.7 https://hardnumbers.dev/pillar/rag2026-09-03T22:26:13+00:00weekly0.7 https://hardnumbers.dev/pillar/inference2026-09-03T22:26:13+00:00weekly0.7 https://hardnumbers.dev/pillar/eval2026-09-03T22:26:13+00:00weekly0.7 https://hardnumbers.dev/pillar/infra2026-09-03T22:26:13+00:00weekly0.7 https://hardnumbers.dev/pillar/devtools2026-09-03T22:26:13+00:00weekly0.7 https://hardnumbers.dev/articles/3-pdf-parsers-tested-docling-vs-marker-vs-paddleocr2026-09-03T09:09:26+00:00 https://hardnumbers.dev/articles/q4-vs-mxfp4-which-quant-is-faster2026-09-03T01:13:52+00:00 https://hardnumbers.dev/articles/4-free-llm-apis-vs-claude-5-coding-tasks-real-data2026-09-02T00:51:53+00:00 https://hardnumbers.dev/articles/gpt-oss-20b-vs-qwen3-14b-vs-mistral-small-24b-a-real-benchmark2026-09-01T10:43:11+00:00 https://hardnumbers.dev/articles/mlx-vs-llama-cpp-on-apple-silicon-same-model-two-engines2026-09-01T11:23:23+00:00 https://hardnumbers.dev/articles/local-llm-tool-calling-for-ai-agents-qwen3-14b-vs-llama-3-2-3b-on-apple-silicon2026-09-01T11:23:23+00:00 https://hardnumbers.dev/articles/qwen3-14b-vs-llama-3-2-3b-on-apple-silicon-m2-an-honest-benc2026-09-01T11:23:23+00:00