Mistral OCR

Pricing : $1/1000 pages, or per 2k pages if “batched”. I’m not sure what batching means in this case: multiple pdfs? Why not split them to halve the cost?

Anyway this looks great at pdf to markdown.

Re: Mistral OCR

#5

Post by newfocogi » Thu, Mar 06, 2025, 6:03 PM UTC

They say: "releasing the API mistral-ocr-latest at 1000 pages / $"

I had to reread that a few times. I assume this means 1000pg/$1 but I'm still not sure about it.

Re: Mistral OCR

#6

Post by pawelduda » Thu, Mar 06, 2025, 6:03 PM UTC

It outperforms the competition significantly AND can extract embedded images from the text. I really like LLMs for OCR more and more. Gemini was already pretty good at it

Re: Mistral OCR

#7

Post by jacksnipe » Thu, Mar 06, 2025, 6:05 PM UTC

vessenes wrote: post #4

Dang. Super fast and significantly more accurate than google, Claude and others. Pricing : $1/1000 pages, or per 2k pages if “batched”. I’m not sure what batching means in this case: multiple pdfs? Why not split them to halve the cost? Anyway this looks great at pdf to markdown.

I would assume this is 1 request containing 2k pages vs N requests whose total pages add up to 1000.

Re: Mistral OCR

#8

Post by abiraja » Thu, Mar 06, 2025, 6:05 PM UTC

vessenes wrote: post #4

Dang. Super fast and significantly more accurate than google, Claude and others. Pricing : $1/1000 pages, or per 2k pages if “batched”. I’m not sure what batching means in this case: multiple pdfs? Why not split them to halve the cost? Anyway this looks great at pdf to markdown.

Batching likely means the response is not real-time. You set up a batch job and they send you the results later.

Re: Mistral OCR

#9

Post by hntiquated » Thu, Mar 06, 2025, 6:06 PM UTC

vessenes wrote: post #3

ChrisArchitect wrote: post #2

[flagged]

No comments there yet - this at the top of the home page, let’s use this one.

[dead]

Re: Mistral OCR

#10

Post by sophiebits » Thu, Mar 06, 2025, 6:06 PM UTC

vessenes wrote: post #4

Dang. Super fast and significantly more accurate than google, Claude and others. Pricing : $1/1000 pages, or per 2k pages if “batched”. I’m not sure what batching means in this case: multiple pdfs? Why not split them to halve the cost? Anyway this looks great at pdf to markdown.

Batched often means a higher latency option (minutes/hours instead of seconds), which providers can schedule more efficiently on their GPUs.