I'm making LLMs work on small and old devices (smolmaxxing)
Spotted 2026-11-13 02:07 (UTC+8) · original event 2026-11-12 · sources: Hacker News · spreading on 1 platformHeat score
Rank score: 7540% — where it sits on the charts
Discussion score: 9040% — how much people are talking about it
Cross-platform score: 3020% — how many platforms it spread to
Composite heat score:72.0rank × 0.4 + discussion × 0.4 + cross-platform × 0.2
Why it matters
Big models need big GPUs; people want usable AI on the old hardware they already own instead of buying new machines.The angle
How tiny models plus aggressive quantization make LLMs run on ancient hardware; what actually works vs what is a stuntSuggested format: Short video / demo · for Developers, AI hobbyists, old-hardware fansEvidence chain
• HN #4, 104 pts/38 cmt (2026-11-13)• GitHub hyperionaidev/smolmaxxing - running LLMs on small/old devices• related: Kindle Paperwhite running SmolLM2 offline (2026-11-07), Reddit iMac G3 LLM post• 2026-11-13 03:07 update: HN #7, 143 pts/58 cmt• 2026-11-13 04:07 update: HN #11, 172 pts/65 cmt• 2026-11-13 05:07 update: HN #10, 197 pts/73 cmt• 2026-11-13 06:07 update: HN #10, 230 pts/84 cmt• 2026-11-13 07:07 update: HN #10, 251 pts/86 cmt• 2026-11-13 08:07 update: HN #10, 268 pts/88 cmt• 2026-11-13 09:07 update: HN #9, 276 pts/92 cmt• 2026-11-13 10:07 update: HN #11, 281 pts/96 cmt• 2026-11-13 11:07 update: HN #12, 288 pts/98 cmt• 2026-11-13 12:07 update: HN #13, 298 pts/102 cmt• 2026-11-13 13:07 update: HN #11, 309 pts/103 cmt• 2026-11-13 14:07 update: HN #14, 323 pts/115 cmt• 2026-11-13 15:07 update: HN #15, 334 pts/121 cmt• 2026-11-13 16:07 update: HN #14, 346 pts/127 cmt• 2026-11-13 17:07 update: HN #13, 360 pts/127 cmt