inbound vs noima

inbound has no write-up on Specifiction yet; its figures are above.

noima extracts requested information from emails, PDFs, scans, and spreadsheets, returning it in structured formats. Users can describe extraction needs in plain language or have AI create classifiers and extractors from sample documents. An AI splitter can divide multi-document PDFs into sections or treat each page separately. Inputs arrive through a workspace email inbox, file upload, or API; supported formats include PDF, EML, HTML, TXT, JPG, PNG, BMP, DOCX, XLSX, XLS, CSV, and Markdown, with a 20 MB limit per file. Results can be exported as Excel, CSV, JSON, or Markdown, or retrieved by API. Small, Balanced, and Pro models use one, two, and three credits per document, respectively. Quotas are shared across an account’s workspaces, and processing stops when the allowance runs out. Failed documents are retried automatically without extra document charges for retries, though an ultimately failed upload still uses one document. The maker says data is hosted in the EU and handled in line with GDPR. Plans include a free one-time allowance and paid tiers starting at €12/mo.