Service Level Agreement

Uptime commitment, response-time targets, and service credit policy.

Version: 1.0  ·  Effective date: 2026-04-22  ·  Applies to: Business and Enterprise plans (Pro on best-effort basis)

Scope. This SLA applies to the Brainiall Services as defined in the Terms of Service. Uptime measurement and service credits apply to the Core Services (chat API, web app, Studio, Academy) for paid tiers. Model-provider–specific downtime upstream of Brainiall's gateway is excluded (see §4); Brainiall's gateway operates automatic failover across providers on best-effort.

1. Uptime Commitment

PlanMonthly uptime targetMax monthly downtime
FreeBest-effort (no commitment)
Pro99.5%~3h 39min
Business99.9%~43min 50sec
Enterprise (signed MSA)99.95% (custom 99.99% available)~21min 54sec

Uptime is measured as the percentage of minutes per calendar month during which the Core Services responded to health-check probes with HTTP 200 within 5 seconds. Live status and historical uptime are published at /status.

2. Incident Severity & Response Times

Severity Definition First response Status update cadence
Sev 1 Complete Service outage affecting all customers ≤ 15 minutes (24/7) Every 30 minutes until resolution
Sev 2 Major feature degraded; significant customer impact ≤ 60 minutes (business hours); ≤ 2h (off-hours) Every 2 hours
Sev 3 Minor degradation; workaround available ≤ 8 business hours Daily until resolution
Sev 4 Cosmetic / low-impact ≤ 2 business days At resolution

Incidents are reported and tracked at /status. Customers on Business and Enterprise plans receive incident notifications by email; Enterprise customers receive Slack-channel or on-call paging on request.

3. Service Credits

If monthly uptime falls below the committed target in §1, qualifying customers may request service credits as follows:

Monthly uptime observedCredit (% of monthly fee)
≥ target0%
target minus 0.1% to target10%
target minus 0.5% to target minus 0.1%25%
Below target minus 0.5%50% (capped at 50% of monthly fee)

Credit requests must be submitted within 30 days of the incident via support@brainiall.com, referencing the incident ID from /status. Credits apply to the next billing cycle and are the sole and exclusive remedy for SLA breaches, in lieu of damages.

4. Exclusions

The following do not count against uptime:

5. Measurement Methodology

Uptime is measured from synthetic probes hitting the /health endpoint every 60 seconds from at least three geographically distributed probe locations. An incident is counted when ≥2 of the 3 probes report failure for ≥2 consecutive minutes. Raw probe data is retained for 6 months and available to Enterprise customers on request.

6. Notification & Escalation

7. Upgrades to this SLA

Brainiall may update this SLA from time to time. Changes that reduce commitments will take effect 30 days after publication and email notification to active customers. Changes that improve commitments apply immediately.

8. Contact