OpenAI API Fast mode (formerly Priority processing) SLAOpenAI · AI models · OpenAI · AI models
AI modelsChecked 15h ago
Promise vs reality
No public SLA, so there is no promise to test. How this is measured
All incidents, weighted: 99.824% · Downtime- Promised
- None
- Observed, 365d
- 100%
- Major incident time
- 0 min
- Credit on first breach
- n/a
Commitments and credits
Docs FAQ: 'Fast mode for GPT-6 Astra does not include a latency SLA. For GPT-5.6 and earlier models, Fast mode and Scale Tier receive the same service-level agreement treatment, and eligible Enterprise agreements may provide service credits when latency targets aren't met.' No numbers are stated on the Fast mode page itself, so tiers is empty (the Scale Tier numbers may apply by reference but that is not stated explicitly).
Terms
- Measured
- Latency targets referenced but not published; SLA terms live in Enterprise agreements.
- How to claim
- Contact your account director
Not covered
- GPT-6 Astra: no latency SLA
- Fine-tuned models and embeddings not supported
- SLA credits only under eligible Enterprise agreements
Evidence: incidents, last 365 days
No incidents on file for this service.
SLA changes
- Jul30OpenAI API Fast mode (formerly Priority processing)Priority processing renamed Fast mode ('Priority processing was renamed Fast mode on July 30, 2026'); service_tier 'priority' still accepted.Source
Other ai models SLAs
ServiceDowntime vs allowed
Amazon BedrockAWS99.9% / 99.178%
Vertex AI (Gemini API on Vertex)Google Cloud99.9% / 100%
Claude API Priority TierAnthropic99.5% / 98.8%
Cohere API (SaaS: Command, Embed, Rerank)Coherenone / 100%
Hugging Face Inference EndpointsHugging Facenone / n/aNo public SLA
OpenAI API Fast mode (formerly Priority processing)OpenAInone / 100%
Promised in the SLAObserved on the vendor's status pagePartial historyShortfall
Summaries of published SLAs; the contract you sign governs. Logos via logo.dev; trademarks belong to their owners.