Baselight @BaselightDB

Explore the new datasets on Baselight: - AI Models Intelligence baselight.app/u/blt/dataset/… - NOAA climate and environmental datasets baselight.app/u/noaa From AI model benchmarks and pricing to climate records going back to 1763 - all queryable as structured data.

We’re tracking 767 AI models across every major lab. What the data says right now: - Top Intelligence Score: Claude Opus 4.8 (61.4) & GPT-5.5 (60.2) - Best value: Gemini 3.5 Flash - 55.3 score at $1.50 prompt / $9 completion (1M tokens) The gap between “best” and “best value” is closing fast. This week we added AI Models Intelligence to Baselight: normalized model pricing, benchmarks, capabilities, endpoints, and performance data from sources including OpenRouter and Artificial Analysis. We also added NOAA NCEI datasets with climate and environmental data from 1763 to present. Baselight now has: - 509,001,889,281 rows - 484,031 tables - 78,773 datasets That’s +4B rows, +7K tables, and +2K datasets this week. Structured data is becoming the intelligence layer for everything - from climate history to the AI model economy.

Want to explore GDELT directly? Dataset link: baselight.app/u/gdelt/datase… Open it, click Ask AI, and ask anything you want to know about global events, media coverage, countries, themes, or trends.

What does the world look like in 2026 - by the numbers? The GDELT Events Dataset has been tracking every major news event globally, every 15 minutes. Here's what 2026 looks like so far: 15.7 million events recorded in just 5 months 62% of events are cooperative - but tone is still negative March 2026 was the most conflict-heavy month of the year 🇺🇸 The US is the most-covered country - 1 in 3 global events 🇮🇷🇮🇱 Iran & Israel both crack the global top 5 Some curiosities regarding Top 10 GDP Nations: 🇺🇸 USA scores the most negative tone in the dataset - every single month. Not one big crisis. Just relentless volume. 🇯🇵 Japan is the only country to go positive - thanks entirely to a single Toyota CEO story in February. 🇫🇷 France improved in May - but only because previous months had kidnappings, firefighter strikes, and a judge scolding Shia LaBeouf. The bar was low. This isn't just data. It's the pulse of the planet. Explore it on Baselight: @gdelt.events_v2

Links to the two new datasets now live on Baselight: CISA Known Exploited Vulnerabilities baselight.app/u/cisa/dataset… GDELT Events v2 baselight.app/u/gdelt/datase…

What do actively exploited cyber vulnerabilities and real-time global news events have in common? They’re both now queryable on Baselight. CISA Known Exploited Vulnerabilities - A real-world cyber risk feed: vulnerabilities the U.S. government confirms are actively exploited in the wild. - Now queryable, joinable with NIST NVD, and ready for exposure analysis, remediation tracking, and security agents. GDELT Events - A global event intelligence feed: news from 100+ languages, scanned every 15 minutes and converted into structured events. - Who did what, where, to whom: with geography, sentiment, and conflict/cooperation signals. Baselight now stands at: 505,217,114,364 rows 476,697 tables 76,488 datasets From cyber risk to global events, Baselight keeps turning the world’s structured data into something humans and agents can query.

Link to the May 22 Baselight Daily Insight - two days before the final - where the odds divergence around Torreense was flagged: baselight.app/daily-insights…

Everyone saw the upset after the final whistle. Baselight had already spotted something unusual days earlier. Baselight Daily Insights are produced by fully autonomous agents that query our verified structured data every day to find anomalies, outliers, inflection points, and signals worth human attention. Today they flagged Torreense’s shock Taça de Portugal final win over Sporting CP as a major outlier: a second-division side winning 2–1 after extra time, despite Sporting being priced around 1.16 and Torreense around 14.5. But the more interesting signal came days earlier. On May 22, Baselight flagged unusual odds divergence around Torreense: Betano priced the win at 32.0, while market consensus was around 17.19 - an 86%+ divergence. Baselight did not “predict the upset”. It surfaced a market anomaly that deserved human review: a possible stale price, model disagreement, or risk miscalculation. That is the goal: autonomous agents turning verified structured data into explainable, auditable signals before they become obvious. Link to the May 22 insight in the comments.

Baselight weekly update 🚀 - after crossing half a trillion rows last week, this week the most interesting signal is breadth. Baselight now includes: . 504,714,444,290 rows (+4B rows this week) . 476,442 tables (+15K tables this week) . 76,438 datasets (+5K datasets this week) We’re now getting close to two nice round milestones: 500K tables 80K datasets That matters because the real value of a universal data platform is not only scale. It’s coverage. More tables means more analytical surfaces. More datasets means more domains, more sources, more context, and more ways for humans and AI agents to connect signals across the world’s structured data. We also added new sources this week, including NIST NVD, bringing cybersecurity vulnerability data into the catalog. But the bigger story is momentum: Baselight is becoming a broader, richer structured data layer every week.

Here is the Baselight query behind the chart, including the weekly CVE severity breakdown from 2024 to 2026: baselight.app/u/pjsousa/quer…

Cybersecurity data is accelerating fast. Baselight now includes NIST NVD data, and the latest numbers are striking: The week of May 11, 2026 saw 1,889 CVEs published in a single week - the highest 7-day total ever recorded in the NIST National Vulnerability Database. And we all know what is driving part of this: GenAI is getting very good at finding vulnerabilities. 📊 Jan avg: ~1,094/week 📊 Feb avg: ~1,198/week 📊 Mar avg: ~1,435/week 📊 Apr avg: ~1,344/week 📊 May so far: ~1,749/week With NVD now in Baselight, CVEs become structured, queryable data that can be correlated with almost 500K tables and 500B+ rows across all knowledge domains. That is the real power: not just tracking vulnerabilities, but connecting them to the wider world of structured data.

Huge week for Baselight - we’ve just crossed a major milestone: 500,142,351,951 rows of data. That’s half a trillion rows now available through Baselight. This week alone we added 17B rows. Baselight now includes: - 500,142,351,951 rows - 461,563 tables - 71,591 datasets And we added two exciting new sources: OpenSanctions: including OFAC, EU Consolidated Sanctions, UN Security Council, and national sanctions lists - plus Politically Exposed Persons (PEPs) registries from 60+ countries. Government of Canada Open Portal: from immigration & permanent resident trends, public health & opioid treatment archives, employment equity & public service workforce stats, to infrastructure data like highways, wastewater plants, and child care maps. Baselight is steadily becoming the place where public data from every domain comes together: searchable, queryable, and ready for humans and AI agents. Half a trillion rows down. Much more to come.

Bookmark Baselight Insights and check it every morning: baselight.app/daily-insights Or add the feed to your agent so it knows what matters every day.

Our Baselight Daily Insights, fully executed by our own team of agents, keep surfacing what matters, every day, grounded in structured and verified data. Recent examples: - ETH closed higher for six days in a row: its longest winning streak for a while, highlighting a short-term momentum shift backed by reproducible queries over live structured data. - Bundesliga Champions League race: three clubs locked on 58 points with two games left, as our agents picked up one of the most intense top-4 battles in European football. This is the direction we're pushing hard with Baselight: AI-generated insights where every claim has a data trail - traceable, reproducible, and grounded in verified structured data. Baselight now has: 482,627,060,226 rows (+2B this week) 459,286 tables (+274) 70,988 datasets (+40) Building the data layer for reliable AI, one verified dataset at a time.

Link to Baselight Daily Insights: baselight.app/daily-insights Bookmark it and check it every morning - or add it to your agent's feed so it can stay on top of what matters.

Baselight update: 240M rows of U.S. mortgage and housing data, +4B total rows, and the launch of Baselight Daily Insights. This week we added Federal Housing Finance Agency data across Federal Home Loan Bank and Fannie Mae / Freddie Mac public-use datasets. That means almost 2 decades of U.S. mortgage and housing data: borrower demographics, credit profiles, loan characteristics, census tract geography, single-family loans, and multifamily properties. You can now explore questions like: - Are young people being priced out? - Where are apartments being built? - Are there lending disparities by race and ethnicity? Baselight now has: 480,465,801,488 rows (+4B this week) 459,012 tables (+1K) 70,948 datasets (+251) We also launched Baselight Daily insights generated by agents that find anomalies, cross-check evidence, rank what matters, and publish the query trail. That means every insight is grounded in structured data, with the underlying queries available to inspect instead of asking you to trust a black-box AI summary. Bookmark it and check it every morning (link in the comments) - or add the feed to your agent so it knows what matters every day.

Baselight Pulse / Daily Insights: baselight.app/daily-insights/

Baselight Pulse is live. Every day, our agent team scans Baselight’s structured data and crafts a fresh set of Daily Insights across: - Soccer - Crypto prices - Stocks The goal is simple: move beyond search and dashboards into proactive intelligence. Instead of waiting for users to ask the right question, Baselight Pulse starts surfacing what may matter: anomalies, divergences, inflection points, unusual patterns, and market or sports moments worth paying attention to. And because these insights are generated from structured data, every insight includes the exact evidence trail behind it: the queries, intermediate results, and data used to reach the conclusion. So this is not just "AI says so." It is auditable, inspectable, and explainable intelligence. We’ve been preparing this over the last couple of weeks, and it feels like another important step towards the kind of data agents we want to build: agents that understand large-scale structured data, detect what changed, and help users discover insights they might not have known to look for. This is still early, and we’d love feedback. Link in the comments. Signal, not noise.

Query it live on Baselight: baselight.ai

Another big week for Baselight. We’ve added the WHO Global Health Observatory: 10.3 million rows of data, across 3,057 indicators, spanning 1932–2030. You can now explore questions like: - How large is the global life expectancy gap? - How does tobacco use prevalence differ by sex worldwide? - What share of TB patients test HIV-positive? Baselight now stands at: 476,066,264,894 rows (+3B) 457,900 tables (+2K) 70,697 datasets (+1K) Approaching half a trillion rows of structured data, all in one place.