Deep Analysis
Hosted DeepSeek R1 reasoning model — reads your document and answers a specific question with explicit step-by-step reasoning. Suited to contracts, legal text, code review, and root-cause analysis from logs. Uses 3 credits per run.
About Deep Analysis
Deep Analysis runs your document through a hosted DeepSeek R1 reasoning model that answers a specific question with explicit step-by-step reasoning. Reach for it when a quick summary won't cut it — contracts, legal text, code review, or root-cause analysis from logs where you need the model to show its work. It's a Pro tool using 3 credits per run.
- Category
- text
- Input
- Accepts: text/plain.
- Output
- Outputs: application/json.
- Cost
- Credit-metered
- Memory
- low
Common uses
- Ask whether a contract clause actually limits liability, and see the reasoning behind the answer
- Walk through a stack of log lines to identify the most likely root cause of an incident
- Have it reason about whether a code snippet has a concurrency bug, step by step
- Analyze a legal notice to determine your obligations and the deadlines that apply
- Pressure-test an argument in a memo by asking the model to find the weakest assumption
- Compare two policy passages and reason out where they conflict
Frequently asked questions
What's different about a reasoning model here?
DeepSeek R1 works through the problem with explicit step-by-step reasoning rather than jumping to a one-line answer, which suits contracts, legal text, code review, and root-cause analysis where the chain of logic matters.
What does it return?
JSON containing the answer along with the reasoning the model used to reach it, given the document and your question.
What do I provide?
Plain text — your document or log excerpt — plus the specific question you want answered.
Is the document sent to a server?
Yes. It's a Pro tool that sends your text to the hosted DeepSeek R1 model for analysis. It's processed for your request and the result is returned to you.
Why does it cost more than the summarize tools?
It uses 3 credits per run because reasoning models do substantially more work per query than a straight summarizer. Use it when the depth of analysis justifies the cost.
Keywords
- analysis
- reasoning
- deepseek
- r1
- contract
- legal
- pro