{"id":"crosby-surgeon","slug":"crosby-surgeon","name":"Surgeon","orgId":"crosby","family":"crosby-surgeon","releaseDate":null,"availability":"announced","openWeights":false,"license":null,"paramsTotal":null,"paramsSource":null,"contextWindow":null,"knowledgeCutoff":null,"modelCardUrl":"https://crosby.ai/blog/how-to-train-your-surgeon","hfRepo":null,"modalities":["text"],"aliases":["Surgeon","crosby surgeon","surgeon (crosby)","crosby-surgeon","surgeon verifier","crosby surgical verifier"],"aliasNote":null,"origin":"tracked","tracked":{"trackedSince":"2026-11-03","announced":"2026-11-03","blurb":"Crosby's in-house verifier model that judges whether a contract redline is surgical (precise, targeted edits that keep the legal substance), returning reasoning plus a yes or no verdict. Post-trained from Qwen-3.5-27B; internal, live only inside Crosby's service for its customers.","api":"none","openWeights":"no","params":{"text":"27B (Qwen-3.5-27B base)","source":"https://crosby.ai/blog/how-to-train-your-surgeon"},"facts":[{"label":"Type","value":"Verifier: reads one redline and outputs a reasoning trace and a yes or no verdict on surgicality"},{"label":"Base model","value":"Qwen-3.5-27B (Crosby's blog)"},{"label":"Training","value":"One epoch of SFT on rejection-sampled teacher traces, then GRPO with a reward gated on a correct verdict times an LLM-judge score of the trace"},{"label":"Training data","value":"Crosby's agent redlines labeled against attorney-sent versions, plus synthetic negatives; validated on 500 redlines relabeled by Crosby attorneys"},{"label":"Use","value":"Offline evaluation, live monitoring, auto-acceptance of surgical redlines and steering Crosby's agentic harness"},{"label":"Availability","value":"Internal/customers only: in production inside Crosby's service; no public weights or API"}],"claims":["Lab claim: 90.6% balanced accuracy on Crosby's internal 500-redline surgicality set, against 84.3% for Gemini 4 Argon, 83.1% for GPT-6 Astra and GPT-6 Sol, and 60.4% for a heuristic baseline.","Lab claim: roughly 10-100x lower cost per million tokens than most closed models in that comparison."],"sources":[{"kind":"lab","title":"How to Train your Surgeon (Crosby blog)","url":"https://crosby.ai/blog/how-to-train-your-surgeon","date":"2026-11-03"},{"kind":"lab","title":"Crosby announcement on X","url":"https://x.com/crosbylegal/status/2117664204568109344","date":"2026-11-03"}],"firstBenchmark":null},"organization":{"id":"crosby","name":"Crosby","slug":"crosby","country":null,"url":"https://crosby.ai","color":"#11110F","mark":"C"},"price":null,"answer":{"word":"Pending benchmarks.","scope":"Announced Nov 3, 2026; no independent results yet.","lead":"As of Nov 22, 2026, Surgeon is pending benchmarks on isitsota: announced Nov 3, 2026, with no independent results on the boards we track yet.","confidence":"low"},"status":"PENDING_BENCHMARKS","verdicts":[]}