Categories › AI & Agents › Models, LLM APIs, routing, serving, training & GPUs
Daily Nerf Watch scores for Claude and GPT models on one private, code-graded suite. Read-only.
Remote server: your requests and data go to tokenusage.dev. Claude Code: claude mcp add --transport http nerf-watch https://tokenusage.dev/mcp/nerfwatch · Claude app: Settings → Connectors → Add custom connector.
The tools this server listed when we connected to it, in its own words.
get_nerfwatch_scores — Latest public Nerf Watch scores, previous runs, changes and baselines for at most 100 models, ordered by provider, name and model id. No argget_nerfwatch_history — Retained public Nerf Watch score history for one model, oldest first, including each point's suite version. The required model id is availabNot tested by KitScout yet. When our harness runs it, pros, cons and the tasks we tried appear here, with the date and evidence.