About RunLog
Make AI choices with evidence rather than guesswork. RunLog helps developers make cost visible, compare model behavior, and add practical checks around agent workflows.
Start with public evidence
Explore recorded task results without signing in, then estimate the cost of your own workload. The public tools are free to explore. Google sign-in is optional.
Explore public benchmarks →Tools you can use today
Public benchmarks show recorded model evidence. Prompt Pricer helps plan costs. Loop-Guard provides local runtime checks, and Ship-Check verifies signed release evidence.
RunLog Systems operates the service. Contact hello@runlog.cc. Google registration is available; paid plans and credit purchases are unavailable. existing access remains available.
The RunLog mark
The RunLog mark represents parallel model paths brought into one place for comparison.
The orange tile is RunLog’s distinctive application mark.
RunLog consulting
Optional support for teams that need a focused agent trace review: a five-day diagnostic with a written memo and team readout. The indicative starting scope is €750; scope, price, and availability are confirmed before engagement.
Ask about consulting →