OpenAI, Anthropic and Google were grappling with service problems Thursday, disrupting several of the tech giants’ massive popular artificial intelligence products.

OpenAI said it was investigating “elevated errors across ChatGPT and Codex,” with the problem ongoing for 20 minutes at the time of a status-page screenshot reviewed by The Post.

Anthropic, meanwhile, reported elevated errors across multiple Claude models and said it had identified the cause and was working on a fix.

The affected models included Mythos/Fable 5.1, Mythos/Fable 5, Opus 5, Opus 4.8 and Opus 4.6, according to Anthropic’s status page.

Anthropic first reported the problem just before 9:30 a.m. Eastern Time. By 10:49 a.m., the company said it was “continuing to work on a fix for this issue.”

Google’s Gemini was also partially degraded for some developer and API users, according to the supplied reporting.

Google AI Studio’s status page said the Gemini API was having problems serving recently created API keys, including keys used through OpenAI-compatible libraries, and that the company was investigating.

The problems hit three major artificial-intelligence platforms at roughly the same time Thursday. The Post has sought comment from OpenAI, Anthropic and Google.