Claude's reliability took a hit in June—two outages dropped it to 90 on our LLM Meter while ChatGPT climbed to 86. When unattended agents run your code, uptime
Anthropic's Claude experienced a decline in reliability during June, with two outages causing its performance on the LLM Meter to drop to 90. Meanwhile, OpenAI's ChatGPT saw an increase, reaching 86 on the same metric. This highlights the growing importance of uptime for AI systems, especially as unattended agents are increasingly used for code execution. AI
IMPACT Highlights the critical need for stable AI performance as unattended agents become more prevalent in operational tasks.