Bank dev tests 1.7B local model offline: all traffic stayed on localhost, emails in 2.5s
Ok_Resolve4741 · reddit · 2026-11-08
A bank employee blocked from putting code or customer data into ChatGPT or Claude tested a local LFM3 1.7B model on his own M4 16GB laptop.
- Privacy first: created a fake customer with an IBAN and balance, had the model draft a late-payment email while logging every connection — all 127 were 127.0.0.1, nothing left the machine.
- Performance: email in 2.5 seconds; a 2-bug Python file fixed in 72 seconds, passing all tests.
- Gotcha: the model hallucinated a "card ending in 4917" from IBAN digits, so outputs still need human review.
Verdict: nowhere near Opus, but good enough for emails, boilerplate, and reading legacy code. He's asking whether anyone has actually gotten local models approved by bank security.
More from coding & agent
- Llama App adds official agent integrations: run Pi and Codex with local models, fully offline — mervenoyann · 2026-11-08
- 10-minute Sherlock Holmes anime made with ComfyUI and Claude Code, files included — shiwano · 2026-11-08
- RustingStrata hits 160 tok/s on RTX 3060 running 35B MoE, ~2x faster than llama.cpp — Serious_Scarcity7818 · 2026-11-08
- Karpathy built a curated video platform in a weekend, filtering 43M HN comments with Claude — The Decoder · 2026-11-08
- Running Multiple LLM Slots on One 7900 XTX Boosts Total Tokens/Sec — c2btw · 2026-11-08
- Codex users report unusually fast burn of 5-hour limits on simple tasks — NeighborhoodPrize493 · 2026-11-08