I haven't been able to parse markdown off of pdf's all day. I'm getting the error with the openai-gpt4o and anthropic-sonnet-3.5. Previously i have not tried sonnet, but gpt-4o has been very reliable for me up until now. Is anyone else having this problem?
Oh good thank you so much. We need the multi modal mode because "accurate" doesn't work well for our documents. I was afraid it was just an issue on my end π