> ## Documentation Index > Fetch the complete documentation index at: https://docs.squasher.ai/llms.txt > Use this file to discover all available pages before exploring further. # Plans and billing > How Squasher packages usage, overages, monthly quotas, and billing behavior. Squasher pricing is designed around a simple rule: teams should understand what they are buying before the invoice lands. ## What is metered Squasher plans package the core surfaces most product teams need in one place: * Error monitoring * Logs * Metrics and traces * Session replay * AI observability * Uptime monitoring and status tooling Every plan includes monthly quotas for the main metered products: * Errors * Telemetry GB for logs, traces, and raw observability data * AI requests * Session replays * Monitor checks ## How monthly quotas work * Quotas reset every billing cycle * Unused quota does **not** roll into the next month * Free plans are hard-capped at the included monthly quota * Paid plans can continue ingesting above the included quota when overages are enabled ## How overages work Paid plans use explicit overage blocks instead of opaque blended pricing. * Error, telemetry GB, AI request, replay, and monitor-check overages are itemized separately * Each meter has a published block size and per-block rate * Usage is rounded up to the next block once you exceed the included monthly quota * The billing dashboard shows both included usage and overage usage for the current cycle Log, trace, and raw AI observability telemetry is billed by ingested GB, not by log-event count. Error events, AI requests, session replays, and monitor checks stay on their own meters because they represent different product value and infrastructure cost. ## What counts as an AI request Squasher meters completed generations as AI requests. Raw telemetry events, spans, and tool calls do not become AI requests unless Squasher records or runs a generation for them. * Observed app or agent generations count * Error triage generations count * Incident summaries count * Auto-fix drafting and related assistant generations count * Tool calls, spans, and other supporting telemetry stay queryable but are not the primary billed unit ## Where to see current plan values Because pricing changes over time, the live source of truth is always: * the public [pricing page](https://squasher.ai/pricing) * the Billing section inside your Squasher dashboard * the billing plans API field `telemetry_gb_limit` for the included telemetry allowance The legacy API field `log_limit` is still returned for older clients, but it now has the same meaning as `telemetry_gb_limit`. ## Retention Retention is feature-specific. Error history is kept longer than high-volume telemetry, while logs, traces, replays, metrics, and AI observations each follow their plan's own retention policy. Enterprise contracts can also carry organization-level retention overrides. The Billing section shows the current plan defaults. Retention changes can take time to physically compact out of long-term storage, but expired data is treated as unavailable once the effective retention window has passed. ## Why the plans are bundled this way Most teams do not want to buy one vendor for errors, another for replay, another for uptime, and a fourth product for AI workflows. Squasher packages these together so incident response stays in one place and budgeting stays understandable. ## Agent handoff When an agent needs current plan data, do not infer prices from old docs or examples: ```text theme={null} Check the current Squasher plan and usage data for project . Prefer the billing dashboard or billing API/CLI output, report included usage and overage status, and do not hardcode plan values. ``` Use `squasher billing ...`, the billing API fields, or MCP `search("billing")` followed by `execute` when live workspace access is available.