Snap decision model picks the right bug file 87.5% of the time, reading 71k tokens in one call
ryanallred · x · 2026-10-13
Early-access testing of the decision model Snap: on 200 real bug reports, Snap read 10 code files (71k tokens) in a single call and picked the file to fix 87.5% of the time in 1.1s; 86.5% with 3x more input.
- The team says Snap outperforms rival Jev within its supported context, and performance keeps improving when the context window is leveraged beyond 32k
- They expect decision models to increasingly do context curation for LLMs — real work on data-intensive workloads rather than plain chat
More from coding & agent
- AI Engineer New York opens with a print daily: 300K-dev telemetry says org, not model — jonmarkgo · 2026-10-13
- One-line decorator turns any combinatorial solver into a differentiable NN layer — PMinervini · 2026-10-13
- Google's enterprise Gemini agent: one memory across Workspace, M365 and Slack, routes work to Claude — BenBajarin · 2026-10-13
- Reverse-Engineering Elgato 4K Capture Card With T3 Code, Latency Down to ~24ms — talkaboutdesign · 2026-10-13
- Air Street Capital's Nathan Benaich: Codex predicted prompts are very good — nathanbenaich · 2026-10-13
- AI bot posting on Reddit for its user sparks debate on where agent approval lines should be — ChatGPTworkBot · 2026-10-13