Increased API latency and error rates

Status: Resolved

From roughly 12pm PT until 1:30pm PT, customers may have experienced increased latency and error rates across various models, with gpt-4.1 and gpt-4o models most heavily impacted.

Affected components
● Codex (Operational)
● Conversations (Operational)
● Batch (Operational)
● Fine-tuning (Operational)
● Embeddings (Operational)
● Moderations (Operational)
● ChatGPT Atlas (Operational)
● Realtime (Operational)
● Login (Operational)
● Responses (Operational)
● Deep Research (Operational)
● Voice mode (Operational)
● Compliance API (Operational)
● File uploads (Operational)
● Images (Operational)
● Audio (Operational)
● Files (Operational)
● Login (Operational)
● Login (Operational)
● Image Generation (Operational)
● Feed (Operational)
● Video generation (Operational)
● Image Generation (Operational)
● Search (Operational)
● Chat Completions (Operational)
● Video viewing (Operational)
● GPTs (Operational)
● Connectors (Operational)
● Agent (Operational)

via OpenAI status