manage_searchDeep Research mode
Long-arc questions. Multi-pass reasoning. ~25 seconds, structured brief out.
Deep Research is the mode you reach for when a single reply will not do the question justice. Iterative retrieval across 40-80 chunks. Three rounds of self-critique. Cross-text reconciliation. Powered by Claude Opus 4.7. The output is a sectioned brief you can re-read for years.
Model: Claude Opus 4.7Latency: ~25 secondsPro tier
Deep Research · Opus 4.7● RUNNING
Round 1 · initial retrieval
Brief · 740 words · 16 sources
Best for
Career trajectory, multi-year dasha analysis, life-arc questions, anything that warrants a brief instead of a chat reply.
What goes in. What comes out.
input
What goes in
- ›A research-grade question (e.g. "what does my chart say about a career change after my 35th birthday?").
- ›Your full chart context — D1, D9, D10, current + upcoming dasha, transits.
- ›Optional: any prior context you want considered (saved readings, journal entries).
output
What comes out
- ›A sectioned written brief — typically 600-1200 words.
- ›A "what the texts agree on / where they differ" reconciliation section.
- ›A bibliography at the foot with every (book, chapter, verse) consulted.
- ›A list of three concrete follow-up questions the brief itself suggests.
Under the hood
The exact pipeline this mode runs.
TriageHaiku 4.5 → routes to research pipeline
Retrieval3 rounds × 15-30 chunks · hybrid + re-query
ReasonerOpus 4.7 with multi-turn self-critique
OutputMarkdown brief · sections + bibliography
Three worked examples
The kind of question this mode is built for, and the kind of reply you would actually get.
Question
I’m 33, in product management, considering switching to my own startup. What does my chart say about the next 5 years?
Reply
TL;DR — your Mercury-Jupiter sub-period from Sep 2027 through Aug 2030 is your strongest entrepreneurial window. The D10 (karma) shows independent enterprise (Sun + Mars in the 7th from Karmesh), but the 11th lord (Jupiter) is currently transiting your 8th — meaning the income from the venture builds slowly before Q2 2026. Suggested action — anchor the launch in Mar-Apr 2027, plan the burn for an 18-month runway, and avoid the Saturn-square-Lagna window Aug 2028 - Feb 2029 for major hiring.
menu_bookBPHS 26.5menu_bookPhaladeepika 12.7menu_bookJaimini 1.2.4menu_bookMC 6.15
Question
My partner and I are considering having a child in 2027. What do our charts together suggest?
Reply
A composite read from both D1 + D7 charts. Strong indicators: your 5th lord (Jupiter) is dignified and your partner’s Saptamesh is in benefic aspect. Caution flags: the proposed conception window (Mar-Jul 2027) overlaps with your partner’s Rahu antardasha — Saravali 38.4 associates this with anxiety in early parenting, not with health issues for the child. Suggested mitigation — Hanuman Chalisa daily through the first trimester, and the Vishnu Sahasranama on Saturdays.
menu_bookBPHS 9.1menu_bookSaravali 38.4menu_bookPhaladeepika 6.12
Question
My Saturn return is approaching. Walk me through what to expect across the full transit.
Reply
Your Saturn return is split into three legs. Leg 1 (Sep 2026 - Apr 2027, Saturn in your 12th): turning inward, expense-and-loss focus, prep work. Leg 2 (Apr 2027 - Nov 2028, Saturn in your Lagna): re-shaping of identity and physical health; this is where the most-visible changes happen. Leg 3 (Nov 2028 - Aug 2029, Saturn in your 2nd from natal Moon): financial restructuring and family-of-origin themes. The classical guidance from Phaladeepika is to anchor in a steady daily practice through the full window — the more disciplined the routine, the cleaner the resolution.
menu_bookBJ 6.14menu_bookPhaladeepika 12.3menu_bookBPHS 41.7
check_circle
When to use Deep Research
- • You have a long-arc question (multi-year, life-stage).
- • You want sources you can quote in a journal entry or share.
- • You are willing to wait 25-30 seconds for a structured brief.
block
When NOT to use Deep Research
- • You have a quick question (use Normal).
- • You are making the actual decision today (use Max Council).
- • You want a chat-style back-and-forth (Normal supports that better).
The other three modes
Same chat. Different depth.
FAQ
How is this different from Normal mode?
Normal mode replies in ~3 seconds with one round of retrieval. Deep Research does three rounds, pulls 40-80 chunks instead of 8-12, runs Opus 4.7 instead of Sonnet 4.6, and writes a structured brief instead of a chat reply.
Why does it take 25 seconds?
Three retrieval rounds and a self-critique pass each round. Each round adds latency but eliminates a class of errors: the second round catches what the first missed, and the third round reconciles disagreement.
Is the brief saved?
Yes. Every Deep Research brief is saved to your journal automatically, and you can re-open, re-share or export it any time.
Which tier do I need?
Deep Research is included in Pro and Max. Free tier users can preview an example brief but cannot run their own.
Try Deep Research on your own chart
Free Kundli first, then ask Panditji anything in Deep Research mode.
