Grep Just Got Smarter: v1.1.0
It's been 10 days since we launched the new Grep. We haven't slowed down. Here's what we've shipped: smarter effort detection, new expert agents, 5 new data integrations, and more.
It's been 10 days since we launched the new Grep and we haven't slowed down. In our latest release we've made Grep way smarter, added dozens more data sources and two new AI Experts! Read more below...
Introducing Effort
You can now choose how much effort Grep puts into answering your question.
We call it "effort" because that's exactly what it is - not just searching deeper, but working harder across the board. More sources consulted. More cross-referencing. More verification. More reviewer passes before you see the final answer.
| Level | What Grep Does | Time | Hours Saved |
|---|---|---|---|
| Low | Quick search across key sources. Answers simple questions fast. | ~30 sec | ~30 min |
| Medium | Searches multiple sources in parallel. Cross-references findings. Checks multiple angles before answering. | 2-3 min | ~3-4 hours |
| High | Comprehensive investigation. Multiple reviewer agents validate the output - checking facts, catching gaps, flagging inconsistencies. Your Expert's work isn't done until it passes review. | 5+ min | ~24+ hours |
High effort means Grep treats your question like an analyst would: gather evidence from multiple sources, verify claims against authoritative data, identify gaps in the research, and synthesize everything into a coherent answer.
Grep even recognizes when your question needs more work.
Ask about competitive landscapes or compare business models, and Grep detects when a quick answer won't cut it. It'll suggest switching to higher effort. You stay in control of the final call.
Built-in Quality Checks - High-effort research gets reviewed before you see it.
When you dial up the effort, multiple reviewer agents validate the output - checking facts, catching gaps, and flagging inconsistencies. It happens automatically as part of the research process.
See What's Happening
Task-Based Research - See exactly what Grep is working on.
Research now runs as structured tasks you can follow in real-time. Watch progress through the new task panel and know exactly when each step completes.
Faster Results - Research streams in smoothly, even on long reports.
The whole experience feels snappier. No more waiting for large responses to load.
More Control, More Formats
- Pause & Steer Follow-ups - Stop a follow-up mid-research, add guidance ("focus on recent developments"), and resume with full context preserved.
- Image Understanding - Drop in a PDF or screenshot and Grep actually sees it. AI-powered visual analysis via Gemini Flash, not just OCR. Floor plans get spatial descriptions. Charts get data interpretation.
- PDF Export - Download your research as a clean, shareable PDF report.
New Expert Agents
GTM Strategy Expert - Your AI sales intelligence analyst. Build ICPs with firmographic and technographic signals. Map entire buying committees with verified contact details via People Data Labs and Pipl. Generate competitive battlecards. Deep-dive account dossiers for strategic deals.
Travel Planning Expert - Compare flights across Google Flights and 500+ airlines via Amadeus. Search hotels, Airbnbs, and VRBOs simultaneously. Calculate total costs including hidden fees. Build multi-city itineraries with day-by-day logistics.
5 New Data Integrations
Middesk - 12 tools for authoritative US business verification. Check tax IDs against IRS records, state registrations across all 50 states, OFAC/BIS sanctions, addresses, liens, bankruptcies, and set up continuous monitoring.
Plus real-time flight and hotel pricing via Amadeus GDS, vacation rentals from Airbnb and VRBO, property listings from Zillow, and corporate officer data from OpenCorporates.
Improved Adverse Media Search
Search multiple news sources at once - Brave News, Google News, ComplyAdvantage, and Opoint - and get deduplicated results automatically. No more checking the same story across four different tools.
And More
- Smoother streaming - Research results render faster with optimized markdown display
- Smarter auto-scroll - Content follows along as research progresses, but stays put when you scroll up to review
- Clearer task updates - See what each research task accomplished instead of technical jargon
- Expanding search box - The input grows as you type longer questions
- Full follow-up questions - See your complete question instead of truncated text
- 20+ bug fixes - Squashed issues across the board for a more reliable experience
That's v1.1.0. We'd love to hear what you think - just reply reach out on X or feedback[at]grep.ai.