Promised uptime vs actual

What cloud, data and AI vendors promise in their SLAs, against what their own status pages show.

Most reliable

observed, last 90 days

1AWSPromises up to 99.999%; no major outage on its status page in 90 days.100%2ConfluentPromises up to 99.999%; no major outage on its status page in 90 days.100%3MongoDBPromises up to 99.995%; no major outage on its status page in 90 days.100%4DigitalOceanPromises up to 99.99%; no major outage on its status page in 90 days.100%5TwilioPromises up to 99.99%; no major outage on its status page in the 56 days on file.100%6VercelPromises up to 99.99%; no major outage on its status page in 90 days.100%7ElasticPromises up to 99.95%; no major outage on its status page in 90 days.100%8PineconePromises up to 99.95%; no major outage on its status page in 90 days.100%9OpenAIPromises up to 99.9%; no major outage on its status page in the 58 days on file.100%10DatadogPromises up to 99.8%; no major outage on its status page in 90 days.100%11NVIDIAPromises up to 99%; no major outage on its status page in 90 days.100%12CoherePublishes no SLA; no major outage on its status page in 90 days.100%13PerplexityPublishes no SLA; no major outage on its status page in 90 days.100%14Scale AIPublishes no SLA; no major outage on its status page in 90 days.100%15AnthropicPromises up to 99.5%; its status page showed 69 min of major outage on Claude API (Standard tier) in 90 days.99.95%16AtlassianPromises up to 99.95%; its status page showed 2 h of major outage in 90 days.99.91%17Google CloudPromises up to 99.99%; its status page showed 3.7 h of major outage on BigQuery in 90 days.99.83%18CursorPublishes no SLA; its status page showed 8.8 h of major outage in 90 days.99.59%19SnowflakePromises up to 99.99%; its status page showed 8.9 h of major outage on Snowflake Postgres (HA instances) in 90 days.99.59%20SupabasePromises up to 99.9%; its status page showed 9.3 h of major outage in 90 days.99.57%21CloudflarePromises up to 100%; its status page showed 7.6 h of major outage on Cloudflare Workers in the 71 days on file.99.55%22GitHubPromises up to 99.9%; its status page showed 27 h of major outage on GitHub Enterprise Cloud in 90 days.98.73%

Every ranking, with the evidence →

Promise vs reality

worst shortfall first

Missed98.733%

GitHub Enterprise Cloud

GitHub promises 99.9%; its status page showed 27 h of major outage in 90 days.

26 incidents on its status page

Missed98.929%

GitHub Actions

GitHub promises 99.9%; its status page showed 23 h of major outage in 90 days.

22 incidents on its status page

Missed99.552%

Cloudflare Workers

Cloudflare promises 99.99%; its status page showed 7.6 h of major outage in the 71 days on file.

27 incidents on its status page

Missed99.589%

Snowflake Service (Data Cloud)

Snowflake promises 99.99%; its status page showed 8.9 h of major outage in 90 days.

25 incidents on its status page

Missed99.589%

Snowflake Postgres (HA instances)

Snowflake promises 99.95%; its status page showed 8.9 h of major outage in 90 days.

25 incidents on its status page

Missed99.57%

Supabase Platform (Postgres, Auth, Data APIs, Storage, Pooler, Realtime, Functions, Management API, Branching, Studio, Logging)

Supabase promises 99.9%; its status page showed 9.3 h of major outage in 90 days.

51 incidents on its status page

Missed99.83%

BigQuery

Google Cloud promises 99.99%; its status page showed 3.7 h of major outage in 90 days.

3 incidents on its status page

Missed99.83%

Cloud SQL

Google Cloud promises 99.99%; its status page showed 3.7 h of major outage in 90 days.

3 incidents on its status page

Missed99.83%

Compute Engine

Google Cloud promises 99.99%; its status page showed 3.7 h of major outage in 90 days.

3 incidents on its status page

Missed99.83%

Cloud Storage

Google Cloud promises 99.9%; its status page showed 3.7 h of major outage in 90 days.

3 incidents on its status page

Missed99.909%

Jira (Cloud)

Atlassian promises 99.95%; its status page showed 2 h of major outage in 90 days.

10 incidents on its status page

Missed99.989%

Cloudflare Business plan (CDN / application services)

Cloudflare promises 100%; its status page showed 11 min of major outage in the 71 days on file.

5 incidents on its status page

Missed99.989%

Cloudflare Enterprise (CDN / application services)

Cloudflare promises 100%; its status page showed 11 min of major outage in the 71 days on file.

5 incidents on its status page

Kept100%

MongoDB Atlas (database clusters)

MongoDB promises 99.995%; no major outage on its status page in 90 days.

8 incidents on its status page

Kept100%

Amazon DynamoDB

AWS promises 99.99%; no major outage on its status page in 90 days.

Kept100%

Amazon EC2

AWS promises 99.99%; no major outage on its status page in 90 days.

Kept100%

Confluent Cloud (Kafka clusters)

Confluent promises 99.99%; no major outage on its status page in 90 days.

15 incidents on its status page

Kept100%

Confluent Cloud for Apache Flink

Confluent promises 99.99%; no major outage on its status page in 90 days.

15 incidents on its status page

Kept100%

Confluent Cloud Fully Managed Connectors

Confluent promises 99.99%; no major outage on its status page in 90 days.

15 incidents on its status page

Kept100%

Confluent Cloud Schema Registry (Stream Governance)

Confluent promises 99.99%; no major outage on its status page in 90 days.

15 incidents on its status page

Kept100%

Droplets (CPU)

DigitalOcean promises 99.99%; no major outage on its status page in 90 days.

Kept100%

Twilio APIs (Twilio-branded APIs)

Twilio promises 99.99%; no major outage on its status page in the 56 days on file.

2 incidents on its status page

Kept100%

Twilio Flex (Critical Scenarios)

Twilio promises 99.99%; no major outage on its status page in the 56 days on file.

3 incidents on its status page

Kept100%

Twilio SendGrid Mail Send API

Twilio promises 99.99%; no major outage on its status page in the 56 days on file.

Kept100%

Vercel platform (content serving)

Vercel promises 99.99%; no major outage on its status page in 90 days.

1 incident on its status page

Kept100%

Volumes Block Storage

DigitalOcean promises 99.99%; no major outage on its status page in 90 days.

Kept100%

WarpStream BYOC

Confluent promises 99.99%; no major outage on its status page in 90 days.

15 incidents on its status page

Kept100%

Amazon RDS (Multi-AZ)

AWS promises 99.95%; no major outage on its status page in 90 days.

Kept100%

App Platform

DigitalOcean promises 99.95%; no major outage on its status page in 90 days.

Kept100%

AWS Lambda

AWS promises 99.95%; no major outage on its status page in 90 days.

Kept100%

Container Registry

DigitalOcean promises 99.95%; no major outage on its status page in 90 days.

Kept100%

DigitalOcean Kubernetes (DOKS) HA control plane

DigitalOcean promises 99.95%; no major outage on its status page in 90 days.

Kept100%

Elastic Cloud Hosted (formerly Elasticsearch Service)

Elastic promises 99.95%; no major outage on its status page in 90 days.

20 incidents on its status page

Kept100%

Elastic Cloud Serverless

Elastic promises 99.95%; no major outage on its status page in 90 days.

20 incidents on its status page

Kept100%

Managed Databases

DigitalOcean promises 99.95%; no major outage on its status page in 90 days.

Kept100%

Pinecone Database (pod-based and serverless)

Pinecone promises 99.95%; no major outage on its status page in 90 days.

9 incidents on its status page

Kept100%

Amazon Bedrock

AWS promises 99.9%; no major outage on its status page in 90 days.

Kept100%

Amazon S3 (Standard)

AWS promises 99.9%; no major outage on its status page in 90 days.

Kept100%

Cloudflare Queues

Cloudflare promises 99.9%; no major outage on its status page in the 71 days on file.

4 incidents on its status page

Kept100%

Cloudflare R2

Cloudflare promises 99.9%; no major outage on its status page in the 71 days on file.

12 incidents on its status page

Kept100%

Confluent Cloud ksqlDB

Confluent promises 99.9%; no major outage on its status page in 90 days.

15 incidents on its status page

Kept100%

GitHub Packages

GitHub promises 99.9%; no major outage on its status page in 90 days.

2 incidents on its status page

Kept100%

OpenAI API Scale Tier

OpenAI promises 99.9%; no major outage on its status page in the 58 days on file.

10 incidents on its status page

Kept100%

Regional Load Balancers

DigitalOcean promises 99.9%; no major outage on its status page in 90 days.

Kept100%

Spaces Object Storage

DigitalOcean promises 99.9%; no major outage on its status page in 90 days.

Kept100%

Vertex AI Platform (training, prediction, pipelines)

Google Cloud promises 99.9%; no major outage on its status page in 90 days.

Kept100%

Datadog Core Services

Datadog promises 99.8%; no major outage on its status page in 90 days.

5 incidents on its status page

Kept99.946%

Claude API Priority Tier

Anthropic promises 99.5%; its status page showed 69 min of major outage in 90 days.

29 incidents on its status page

Kept100%

Gemini API on Vertex AI

Google Cloud promises 99.5%; no major outage on its status page in 90 days.

Kept100%

NVIDIA Cloud Functions (NVCF)

NVIDIA promises 99%; no major outage on its status page in 90 days.

1 incident on its status page

Every service and its SLA →

SLA changes

what vendors changed

Number changedOct 27, 2026

Azure OpenAI in Foundry Models

The SLA on file is now Service Level Agreement for Microsoft Online Services, October 1, 2026, replacing the September 1, 2026 edition.

SLA page foundUnverifiedOct 23, 2026

Google

Possible SLA page found: Google Workspace Service Level Agreement. Not yet reviewed.

SLA page changedUnverifiedOct 23, 2026

Stripe API and payments services

The SLA page text changed (30 sentences added, 30 removed); not yet reviewed.

SLA page changedUnverifiedOct 23, 2026

Supabase Platform (Postgres, Auth, Data APIs, Storage, Pooler, Realtime, Functions, Management API, Branching, Studio, Logging)

The SLA page text changed (1 sentence added, 1 removed); it still states 99.9%.

SLA page changedUnverifiedOct 23, 2026

Snowflake Service (Data Cloud)

The SLA page text changed (14 sentences added, 3 removed); it still states 99.99%, 99.95% and 99.9%.

SLA page changedUnverifiedOct 23, 2026

OpenAI API Fast mode (formerly Priority processing)

The SLA page text changed (9 sentences added, 11 removed); not yet reviewed.

SLA page changedUnverifiedOct 23, 2026

Okta Workforce Identity Cloud

The SLA page text changed (18 sentences added, 17 removed); not yet reviewed.

SLA page changedUnverifiedOct 23, 2026

MongoDB Atlas (database clusters)

The SLA page text changed (6 sentences added, 1 removed); it still states 99.995%.

SLA page changedUnverifiedOct 23, 2026

Azure Virtual Machines

The SLA page text changed (6 sentences added, 2 removed); not yet reviewed.

SLA page changedUnverifiedOct 23, 2026

Mistral Enterprise

The SLA page text changed (7 sentences added, 3 removed); not yet reviewed.

SLA page changedUnverifiedOct 23, 2026

Mistral AI Studio / La Plateforme API

The SLA page text changed (4 sentences added, 3 removed); not yet reviewed.

SLA page changedUnverifiedOct 23, 2026

Gemini API on Vertex AI

The SLA page text changed (5 sentences added, 5 removed); it still states 99.5% and 99%.

Every change, going back →