Code model eval: Rust hits 95-100% accuracy as 13 languages see big gains
shuding · x · 2026-11-18
Vercel co-founder shuding shared code-model accuracy results showing 13 languages improved by more than 5% this round: cmake and rust now at 95100%; hlsl, docker, powershell, make, coffeescript, wasm and wgsl at 9095%; haxe, bat, objective-c, less and nim at 8090%. The jump suggests code models are rapidly closing the gap on long-tail languages and DSLs like shaders and shell scripts.
More from coding & agent
- GLM-5.3 pricing called absurdly cheap: $10 plan reportedly unlocks $500 in usage — rez0__ · 2026-11-18
- LLM directly edits π0.5 VLA weights, no fine-tuning, lifting robot task success from 30% to 90% — Rewkang · 2026-11-18
- An entire photorealistic video rendered purely with code — no video model, no Blender — JeremyNguyenPhD · 2026-11-18
- Pawel Huryn open-sources 45 agent skills distilled from shipping three live products — PawelHuryn · 2026-11-18
- Splash 1.7.0 released: faster decoding for coding agents on Apple Silicon — wojtek15 · 2026-11-18
- Law school AI course wraps: 6 student teams ship working legal AI prototypes — jkubicki · 2026-11-18