Need a hand?
Try the getting-started guide or choose help below. You do not need an account to use the public tools or contact support.
Benchmark or catalog did not load?
Use the in-page retry where available. Previously loaded benchmark results remain visible if a refresh fails. Check Status for the public dataset and price catalog; a successful data check does not guarantee model-provider execution.
Pricer can use its labeled bundled price snapshot when live retrieval fails. Shared settings use their saved price snapshot so a later catalog update does not silently change a replayed estimate.
Estimate sharing and exports
Share settings to reproduce numeric token, media and history assumptions without sharing raw text. Copy report exports the current model selection. Estimates are not measured responses; compare all assumptions before using them to plan spending.
Existing-account information
Accounts, Workspace and private comparisons are legacy previews under review. If you see “Checking sign-in”, “Identity service unavailable” or an unavailable catalog, use existing account access, retry after checking status, or contact support. These failures should not prevent you from using public tools.
Saved reports and their expiry are available in legacy Workspace; recent private comparisons are in Studio history. Download important reports before their displayed expiry.
Contact support
Email hello@runlog.cc with the page URL, time, browser and exact visible error. Do not include passwords, API keys, session cookies or confidential prompt text.