10 July 2026

 

AI tools status: No major engine releases in the past 24 hours, but ongoing expansion of AI search, multimodal input, and long-form writing tools continues to support genealogy workflows.

In other words, today is a “steady progress” day rather than a “major launch” day, but the existing ecosystem of search‑integrated AI, long‑form narrative engines, and faster hardware is directly usable by working genealogists.

    New AI releases: GPT‑5.6 Sol/Terra/Luna, Claude Sonnet 5, Gemini 3.1 Pro/Flash, refreshed Perplexity models, and new open‑weight options expand reasoning, speed, and multimodal support.

    Research workflow impact: Cheaper frontier reasoning and emerging agent/workspace tools (ChatGPT Work, Claude recap/focus) make long‑running genealogy projects easier to manage and analyze

Below are concrete, immediately testable examples, framed for a working genealogist or family history blogger. Each can be done with general‑purpose AI models plus your existing tools (Zotero, RootsMagic, FamilySearch, Ancestry, etc.).  

Genealogy micro‑workflows: Over 20 concrete workflows show how to use these tools for locality studies, probate triage, city directories, multimodal transcription, FAN‑club clustering, and public ancestor sites.  

Practical AI uses in genealogy (at least 20)

Research and record analysis

1. Summarizing long pension files
   Upload or copy sections of Civil War or Revolutionary War pension files and have AI produce section‑by‑section summaries with an events timeline and person index.

2. Extracting entities from deeds and land records
   Paste a transcribed deed; ask AI to list all grantors, grantees, neighbors, landmarks, and dates, then output a table you can import into Zotero or a spreadsheet.

3. Jurisdiction tracing for Indian Territory and Oklahoma
   Provide a short description of a location and time (e.g., “near Muskogee, 1895”) and have AI outline likely jurisdictions and record sets (federal, territorial, tribal, county) to check.

4. Probate packet “case file” creation
   Feed AI your notes from a multi‑document probate packet and ask it to synthesize a narrative: who died, heirs, property list, disputes, and key evidence for each relationship.

5. Historical context for tribal rolls
   Give AI the name of a roll (e.g., Dawes, Henderson, Muster rolls) and date range, plus a family’s appearance, and ask for a concise context paragraph explaining what the roll captured and its limitations.

6. Census comparison across decades
   Paste extracted census data for a household from multiple years and ask AI to highlight changes in age, occupation, neighbors, and inferred migration, then flag contradictions needing further investigation.

7. Cluster research planning
   List several FAN (Friends, Associates, Neighbors) names; have AI propose research questions and sources for each cluster, turning your one‑family project into a structured neighborhood study plan.

8. Name and identity disambiguation
   Describe competing hypotheses for two individuals with the same name in the same county and feed AI the evidence bullets; ask it to lay out pros/cons for each identity and suggest targeted follow‑up searches.

Transcription, translation, and cleanup

9. Transcribing difficult handwriting from typed notes
   After you manually rough‑transcribe a few lines of hard‑to‑read script, ask AI to standardize spelling, expand abbreviations, and flag uncertain words with suggested alternatives.

10. Translating foreign‑language records
    Paste snippets from German, Spanish, French, or Scandinavian parish registers and have AI translate into English, preserving key genealogical terms (e.g., baptism, marriage, burial, witnesses).

11. Normalizing place names over time
    Give AI several spellings of a town or county from different documents and ask it to identify the modern standardized form, plus alternate historical spellings for your research notes.

12. Cleaning OCR output from printed histories
    Run a county history through OCR, then ask AI to remove obvious OCR artifacts, break paragraphs at logical points, and mark person/place names for easier excerpting into your research log.

Writing, blogging, and narrative construction

13. Drafting ancestor profile posts from research notes
    Paste your bullet‑point research notes on an ancestor; have AI produce a concise blog‑style profile with sections for early life, migration, records consulted, and unresolved questions.

14. Turning research logs into story outlines
    Give AI a chronological research log; ask it to transform the log into an outline for a narrative article or booklet, highlighting turning points and key evidence moments.finance.yahoo

15. Creating “case study” articles for society newsletters
    Provide anonymized case details (problem, sources, reasoning, conclusion); have AI draft a short educational case study suitable for a society newsletter, which you then fact‑check and personalize.

16. Generating titles and meta‑descriptions for blog posts
    After drafting a post, ask AI for 5–10 alternate titles and a 1‑sentence meta description that emphasize genealogy keywords and clarity.

17. Structuring book‑length family histories
    Feed AI a list of families, places, and time periods; have it propose a chapter structure and section headings for a family history book or multi‑part blog series.finance.yahoo

18. Summarizing webinars and conference sessions
    Paste your session notes; ask AI to summarize the key takeaways, then propose 3–5 concrete actions for your own research and teaching practice.

Teaching, workshops, and pedagogy

19. Creating lesson plans for genealogy classes
    Specify level (beginner/intermediate/advanced), topic (e.g., “probate records”), time (60 minutes), and audience; have AI generate learning objectives, a step‑by‑step lesson plan, and practice activities using sample documents.

20. Designing handouts and checklists
    Ask AI to turn your outline of a topic (e.g., “Using land records in Oklahoma research”) into a one‑page checklist or reference handout for students, ready to format in Word or Canva.

21. Building scenario‑based exercises
    Describe a fabricated but realistic research puzzle; have AI generate questions students must answer, plus a suggested sequence of records to consult and hints you can reveal during class.

22. Creating slide outlines with speaker notes
    Provide your workshop goals and main points; ask AI to propose slide titles and bullet points, plus brief speaker notes you can adapt.

Data management, workflows, and Zotero integration

23. Designing Zotero tag and collection schemes
    Describe your current Zotero use; have AI propose a consistent tagging and collection naming system for record types, localities, and evidence status.

24. Drafting Better Notes templates
    Ask AI to help design structured note templates for deeds, probate, tribal enrollment cards, or military records, including fields for citation, informant, reliability, and research questions.

25. Creating batch‑import schemas
    Describe how your CSV export from Ancestry or FamilySearch looks; have AI suggest column mappings, field names, and normalization rules to feed into your Python batch‑import scripts for Zotero or RootsMagic.

26. Automating research log prompts
    Use AI to generate daily or weekly research prompts based on current projects (“3 next steps for the Creek family in Indian Territory”), which you can paste into your task manager or calendar.

27. Designing a multi‑monitor research workflow
    Explain your hardware and tools; have AI propose window layouts and task groupings (search, transcription, notes, AI assistant) to minimize context‑switching during deep research sessions.

DNA and correlation

28. Summarizing DNA match lists and clusters
    Paste anonymized information about matches (shared cM, notes, trees); ask AI to group them into likely clusters and suggest which ancestral couples they may connect to, for your own interpretation.

29. Drafting plain‑language explanations of DNA evidence
    After you complete your own analysis, feed AI a technical summary; ask it to rewrite in accessible language for non‑specialist family members while preserving the logic and caveats.

30. Integrating documentary and DNA evidence in proof notes
    Provide bullet points of both record‑based and DNA‑based arguments; ask AI to draft a coherent, structured proof paragraph that you then edit and verify against standards.

Plug‑and‑play AI micro‑workflows

Below are at least twenty concrete, genealogy‑specific micro‑workflows, each explicitly tied to one of this week’s releases or model‑line changes. All assume current access through ChatGPT, Claude, Gemini, Perplexity, or an open‑weight stack.perplexity+3

1. Sol‑powered locality evidence synthesis (GPT‑5.6 Sol)

   • Use Sol in ChatGPT to ingest a short locality file (summary of a county in Oklahoma Territory or Indian Territory), plus 5–10 key records, and ask it to articulate 3–5 competing hypotheses for an ancestor’s residence or tribal affiliation, with strengths/weaknesses for each.openai+1

2. Probate packet triage at scale (GPT‑5.6 Terra)

   • Run Terra over a batch of lightly transcribed probate documents to: identify parties, relationships, property descriptions, and jurisdictional clues; then have Terra create a table you can paste into Zotero or Airtable.engadget+2

3. Bulk city‑directory extraction (GPT‑5.6 Luna)

   • Use Luna for fast, low‑cost passes over pages of city directories: extract name, address, occupation, and year for one surname cluster, then export the results to CSV for timeline building.engadget+2

4. Model‑migration checklist for research logs (Retirement of o3, GPT‑4.5)

   • Ask ChatGPT (with GPT‑5.6) to review your past prompt templates or research‑log instructions that mention o3 or GPT‑4.5 and propose updated wording optimized for Sol/Terra/Luna, including which variant to use for each task type.openai

5. Project‑style “case file” assistant (ChatGPT Work)

   • Spin up ChatGPT Work for one stubborn research problem (e.g., identifying a Cherokee enrollee’s parents across Dawes, census, and land records), then keep all questions, draft proof summaries, and source analyses inside that one agent over several days.openai+2

6. Mini research site for a single ancestor (ChatGPT Sites)

   • Use ChatGPT Sites to generate a simple public page for one ancestor or one family line, populated with a timeline, source list, and curated narrative, then iteratively refine the content in ChatGPT before publishing the site for cousins to review.releasebot

7. Monthly research recap and focus planning (Claude recap/focus)

   • Let Claude’s new monthly recap feature summarize what you’ve worked on (e.g., all sessions related to Creek Nation land claims), then ask it to propose a 2‑week focus plan, including quiet hours and break reminders tied to your most demanding analysis work.releasebot+1

8. Session‑level “what changed?” overview (Claude recap)

   • At the end of a day with multiple Claude chats, ask the recap tool to list the three most significant research insights or negative findings you produced, and paste that into your research log as a daily summary.releasebot+1

9. Long‑context correlation of land, tax, and census (Sol or Claude Sonnet 5)

   • Use a high‑reasoning model (Sol in ChatGPT or Claude Sonnet 5 via an API/tool) to load a multi‑source summary: land patents, tax lists, and census entries for a surname cluster, and ask it to identify possible same‑person groupings, conflicts, and gaps.llm-stats+1

10. Photo‑plus‑text “story seeds” (Gemini 3.1 Pro multimodal)

    • With Gemini 3.1 Pro, upload a family photo plus a brief note (e.g., “family in Seminole County, 1910s”) and ask for: clothing/setting clues, plausible local events, and a short outline for a narrative you can then verify against records.chroniclemakersyoutubeperplexity

11. Handwritten pension transcription (Gemini 3.1 + Flash TTS)

    • Feed images of Civil War pension files into Gemini for text extraction, then use Flash TTS to listen through the transcription while following along in the original images, catching errors by ear.youtubedenyseallen.substack

12. Perplexity “reasoning toggle” for tricky land chains (Sonar + GPT‑5.2)

    • In Perplexity, run a quick search with Sonar to gather background on a county’s land‑record system; then enable the reasoning toggle with GPT‑5.2 to analyze a chain of title abstract and explain the sequence of conveyances in plain English.perplexity+1

13. Cross‑platform “where do these records live?” (Perplexity with Gemini 3.1 Pro)

    • Use Perplexity’s multi‑model environment to ask for current online and offline locations of specific record types (e.g., Creek Nation citizenship rolls, Oklahoma Territory land grants), with cited links and repository names.perplexity+2

14. Case‑study microsite using ChatGPT Sites + Perplexity citations

    • Draft a proof argument in ChatGPT (Sol or Terra), then ask Perplexity to generate a clean, cited “sources consulted” section with updated URLs and repository names, and embed that into a ChatGPT‑generated site about the case.perplexity+1

15. Open‑weight image model for map annotations (DiffusionGemma 26B‑A4B)

    • Use an open‑weight diffusion model like DiffusionGemma to generate simple illustrative map‑style images (not records) showing migration paths, then annotate them manually in a graphics tool for use in blog posts or talks.llm-stats+1

16. Open‑weight text model for offline surname‑cluster notes (GLM‑5.2 or Seed 2.1)

    • Deploy GLM‑5.2 or Seed 2.1 locally to help you summarize offline research notes—court minutes, local histories, transcribed interviews—without sending sensitive data to the cloud.llm-stats+1

17. Iterative “research like a pro” agent loop (ChatGPT Work + Perplexity + Claude)

    • Use ChatGPT Work for planning (checklists, next steps), Perplexity for live catalog and locality research, and Claude for drafting narrative sections, aligning with the emerging multi‑tool genealogy workflows described in recent guides.perplexity+3

18. Revising your AI research‑plan templates (Sol/Terra + Claude focus tools)

    • Ask Sol or Terra to review your existing AI‑assisted research‑plan template and suggest improvements that exploit higher reasoning, then use Claude’s focus settings to time‑box actual execution sessions (e.g., 45‑minute “land plan” blocks).engadget+2

19. Timeline‑driven story scaffolding (Claude Sonnet 5 + Gemini)

    • After transcribing records in Gemini, feed the cleaned text into Claude Sonnet 5 and ask for a neutral timeline and section outline, explicitly separating “facts from records” from “interpretive context” you’ll add later.llm-stats+1youtube

20. Multi‑county FAN‑club clustering (Sol or Luna + Perplexity)

    • Use Luna for rapid passes over index data (extract names, associates, locations), then hand the structured output to Sol or a Perplexity reasoning model to identify recurring neighbors, witnesses, and tribal associates across different jurisdictions.perplexity+2

21. Monthly “portfolio” review of active research cases (Claude recap + ChatGPT Work)

    • At month‑end, let Claude recap your case‑related chats, then ask ChatGPT Work to build a single overview document listing each open research question, current status, and next record type to pursue.releasebot+1

22. “52 ancestors” micro‑stories site (ChatGPT Sites + Gemini images)

    • Combine ChatGPT Sites with Gemini’s ability to describe and contextualize existing family photos, creating a simple site where each ancestor gets a short, documented vignette and at least one curated image.