Support for PDFs, DOCX, EPUB, text files, and more.
Ask questions and instantly get complete answers using state-of-the-art AI.
Every response includes citations from the documents.
Our AI intelligently breaks down documents and creates multiple search indexes for more accurate, thorough analysis.
Our AI intelligently breaks down documents and creates multiple search indexes for more accurate and thorough analysis.
Visualize complex analysis with ease—AI works on every document and returns results in a structured table.
Perfect for users focused on data privacy and security.
Select any set of documents to ask questions.
Some documents are LONG. Tell the AI to look in specific sections to get more focused answers.
Reuse simple or complex prompts with any chat.
1 cloud chat
2 files per chat
100 pages/mo for Socrates Deep Dive
2 rows (files) in Table AI
50 Prompt Loop Segments/mo
20 messages (GPT 4o) per month
OCR not included
Sharing not included
Force OCR not included
Unlimited Auto-OCR for Cloud Chats
Unlimited files per chat
1,000 pages/mo for Socrates Deep Dive
20 rows (files) in Table AI
750 Prompt Loop Segments/mo
Unlimited messages (GPT 4o) per month
Auto-OCR
Sharing not included
Force OCR not included
Unlimited Auto-OCR for Cloud Chats
Unlimited files per chat
3,000 pages/mo for Socrates Deep Dive
50 rows (files) in Table AI
2,000 Prompt Loop Segments/mo
Unlimited messages (GPT 4o) per month
Auto-OCR
Share up to 1 Chat
Force OCR not included
Pay As You Go purchases do not expire and will be used only if the monthly limit is reached.
Education discounts available.Contact Support.
Yes, there is a free plan that lets you play with all of the features of the product. Using Local models like Meta Llama is completely free.
Socrates processes your files in multiple ways, with each pass generating additional memories from which Socrates AI can draw insights. You can learn more here.
Socrates supports over 60 languages.
Your documents stay local on Mac or Windows apps, with options to use local LLMs like Llama 3 or send encrypted data to Microsoft Azure OpenAI. Microsoft doesn’t train on this data per their data processing agreement.
Learn more about our Socrates' AI Data Privacy Guide or Microsoft's Azure OpenAI Data Privacy policy.
If you have an nVidia GPU, you can install CUDA, which will provide the speediest local LLM experience. If you have any other GPU, Socrates automatically uses your GPU to speed up the local LLM using Vulkan.
Chats can be trained on an unlimited number of files. For cloud chats, each individual file has a max of 10 million characters or max of 150 MB. Each cloud chat supports a max of 100 million characters.
For local chats, there is no size restriction other than your computer's storage.
Can’t find the answer you’re looking for? Reach out to our customer support team.