Each page here takes one managed AI service or vendor product and sets out what its publisher states about price, data use, model lifecycle and exit, next to the direct model APIs and the option of running an open-weight model yourself. Every fact links to its source and carries the date it was read; model prices move often, so the date matters more here than anywhere else in the library.
| Vendor | Page | What the page answers | Facts checked |
|---|---|---|---|
| Amazon Bedrock | AWS Bedrock alternatives: other managed clouds, direct model APIs or a self-hosted LLM | What Amazon Bedrock publishes about prices, data use and model retirement, next to Azure OpenAI, Vertex AI, direct model APIs and a self-hosted LLM. | Sep 23, 2026 |
| Azure OpenAI | Azure OpenAI alternative: when the retirement clock and the PTU bill decide for you | Azure OpenAI: GPT-5.6-terra at $2.00 and $12.00 per million short-context tokens, PTUs by the hour, models retired after 18 months, three alternatives. | Sep 23, 2026 |
| Amazon Textract | AWS Textract alternative: rent another OCR API or own the document pipeline | Amazon Textract: $0.0015 a page for text, $0.05 for forms. Data terms, regions and languages next to Google Document AI, Azure and your own pipeline. | Sep 23, 2026 |
| Fin | Fin (formerly Intercom) alternatives: what the Salesforce deal changes | Fin (formerly Intercom) joined Salesforce on 10 September 2026. What the $0.99 outcome price counts, what Intercom seats cost, when an own agent pays off. | Sep 23, 2026 |