LLM API Change Monitor
Provider changes break production quietly: a model reaches end of life, a price moves, a rate limit tightens, a clause in the terms changes. This module watches the providers you choose and tells you what changed, what it affects and by when you need to act.
What it covers
- Model releases, aliases and version pinning changes
- Deprecation and retirement notices, with dates
- Pricing changes: per-token, cached, batch and fine-tuning prices
- Rate limits, quotas and context-window changes
- API changes: new, renamed or removed parameters and endpoints
- Terms of service, data-use and retention policy changes
You select the providers and models you use at onboarding. Typical coverage includes the major hosted model APIs and the main cloud model platforms; the list is agreed with you in writing.
What you receive
- Alerts by email and in the portal when a relevant change is detected. Each states what changed, the effective date, the source link and the suggested action, if any.
- Weekly summary with open deadlines sorted by date.
- Change log in the portal, searchable by provider and model.
Illustrative example
Deprecation notice · Example Provider · model
example-large-2025-01Retirement date: 90 days from notice. Replacement named by the provider:example-large-2026-03. Price per million tokens changes from 3.00 to 2.50 input and 15.00 to 12.00 output. Context window: unchanged. Behaviour differences listed by the provider: tool-call format. Source: provider changelog (link). Action: test the replacement model before the retirement date.
This example is fictitious and only shows the format.
How it works and limits
Sources are public: provider documentation, changelogs, pricing pages, status and policy pages. Detection is automated and reviewed before sending. Alerts are normally sent within one business day of publication, but no delivery time is guaranteed. We do not have private information from providers and we do not test your integration. See the Terms of Service.