Something that didn’t fit was discovered by NASA’s Fermi Gamma-ray Space Telescope early on July 2, 2025. Somewhere in the direction of the constellation Scutum, close to the crowded, dusty central plane of the Milky Way, a gamma-ray burst—the most powerful class of explosion in the known universe—had erupted. That was not unusual in and of itself. Roughly once a day, gamma-ray bursts are detected, and they can come from any direction in the sky at any time. This one was unique in that it continued. The duration of most bursts ranges from a few milliseconds to a few minutes. This one, known as GRB 250702B, generated a first gamma-ray wave that persisted for at least seven hours. When it finally subsided, the event had triggered Fermi’s instruments several times over the course of three hours, been picked up by five other spacecraft on two continents as well as a space station, and produced a deluge of data that astronomers are still analyzing months later.
Eliza Neights of George Washington University and NASA’s Goddard Space Flight Center stated, “This is definitely an outburst unlike any other we’ve seen in the past 50 years.” That is a cautious statement from a cautious scientist, but it carries a lot of weight. Since the phenomenon was first discovered in 1973, about 15,000 gamma-ray bursts have been observed; none have lasted as long as this one. Few even come close. It is both thrilling and genuinely unsettling for the theorists who dedicate their careers to creating models of how these explosions operate because it falls into a category all by itself.
Black Holes Eating Stars — Key Discoveries 2025
| Primary event: GRB 250702B | Discovered July 2, 2025 — a gamma-ray burst (GRB) of unprecedented duration; initial gamma-ray wave lasted at least 7 hours, nearly twice the longest GRB previously recorded; of ~15,000 GRBs observed since 1973, none are as long |
| Location of GRB 250702B | Constellation Scutum, near the crowded plane of the Milky Way — precise location obtained July 3 via NASA’s Neil Gehrels Swift Observatory X-Ray Telescope |
| Leading explanation | Black hole consuming a star — two competing theories: (1) an intermediate-mass black hole (~thousands of solar masses) shredding a passing star via tidal disruption; (2) a stellar-mass black hole merging with and consuming a stellar companion from within |
| Detection network | NASA’s Fermi Gamma-ray Space Telescope (Gamma-ray Burst Monitor), Neil Gehrels Swift Observatory (Burst Alert Telescope), NASA’s Wind mission (Russian Konus instrument), NASA’s Psyche spacecraft, Japan’s MAXI on the ISS, China’s Einstein Probe — combined data required because no single instrument could fully observe the duration |
| Lead researcher | Eliza Neights, George Washington University / NASA Goddard Space Flight Center — findings shared at the American Astronomical Society’s High Energy Astrophysics Division meeting, St. Louis, October 2025 |
| Record-setting flare (separate event) | Peak brightness: 10 trillion times brighter than the sun — tidal disruption event involving a star at least 30 times the mass of the sun, shredded by a black hole ~300 million times the sun’s mass, located ~11 billion light-years from Earth; announced November 2025 |
| Three separate NASA events (June 2025) | NASA reported three extreme tidal disruption events involving supermassive black holes consuming massive stars — each released more energy than 100 supernovae combined |
| Neutron star simulation (June 2025) | Caltech team led by Elias Most published simulations of neutron star’s final milliseconds before black hole consumption — predicting violent starquakes, Alfvén waves, and a final fast radio burst (FRB) detectable by future radio telescope arrays including planned 2,000-dish network in Nevada |
| What is a tidal disruption event (TDE) | When a star passes within a black hole’s tidal radius, gravitational forces stretch it into a stream of gas — the material spirals inward, heating and releasing enormous energy as it approaches the event horizon |
| Why GRB 250702B matters | Standard GRB models (neutron star mergers or massive star collapse) cannot explain jets firing for days; the event likely heralds a new class of stellar explosion — multiple competing papers published or accepted; more in preparation |
The two main explanations for gamma-ray bursts are the collapse of a massive star when its nuclear fuel runs out and the merger of two dense neutron stars. Black holes are produced by both. Both produce particle jets that shoot outward at almost the speed of light, producing the gamma-ray flash that is picked up by observatories. The issue is that those jets cannot be sustained for days by either of these mechanisms. The astrophysics community is unable to agree on a single explanation for a burst that fires nonstop for seven hours and exhibits additional activity the day before its peak detection. Researchers concur that the most plausible explanation is that a star was eaten by a black hole, but they are still at odds over the precise mechanism. According to one theory, a star that strayed too close could be destroyed by an intermediate-mass black hole that weighs thousands of times the mass of the sun. Another group suggests a smaller black hole that consumed its stellar companion from the inside out after merging with it. By all standards, both are unusual occurrences.
The fact that the detection itself required the cooperation of five different spacecraft speaks volumes about the scope of what transpired. This duration of observation was beyond the capabilities of any single instrument in orbit. According to Louisiana State University astrophysicist Eric Burns, “we could only comprehend this event through the combined power of instruments on multiple spacecraft.” In actuality, this required combining data from China’s Einstein Probe, which was instrumental in detecting X-ray activity from the burst a full day prior to the main event, Japan’s all-sky X-ray monitor aboard the International Space Station, NASA’s Fermi and Swift observatories, instruments aboard NASA’s Wind mission, and the Psyche spacecraft currently en route to an asteroid. One of the things that makes 250702B so unique and valuable is that day-earlier signal. It implies that something was going on prior to the spectacular gamma-ray outburst.
This was not an isolated incident. A different record was set in November 2025 when a supermassive black hole about 300 million times the mass of the sun, located about 11 billion light-years from Earth, destroyed a star at least 30 times the mass of our sun. At its height, the flare was ten trillion times brighter than the sun. Before it reached our telescopes, the light from that devastation had been traveling across space for longer than Earth had existed. We just had the right instruments pointed in the right direction at the right time to catch it. NASA revealed three distinct cases of supermassive black holes devouring massive stars earlier that year in June, each of which released more energy than 100 supernovae put together.
The fact that all of this is happening within the same 12-month period is somewhat startling. Black holes are among the most theoretically rich objects in astrophysics; they are philosophically unsettling, mathematically beautiful, and difficult to observe. The behavior of black holes consuming stars was not observed in anything approaching real time for the majority of the history of modern astronomy; instead, it was inferred from models and afterglows. This is changing as a result of enhanced detection algorithms, coordinated multi-spacecraft observation networks, and better instruments. There is a sense that the field is about to enter a phase where the once purely theoretical events are gradually becoming things that can be observed and compared against the models, as researchers in St. Louis, Louisiana, Washington, and Greenbelt watch the data pour in from multiple satellites at once in order to understand what they’re seeing.
There will be more GRBs like GRB 250702B. Throughout the history of the universe, there are probably more occurrences of this length and peculiarity that were just too long-lasting or sensitive for earlier instruments to detect. Now, astronomers are aware of what to look for and have enough networked hardware in orbit to detect it should it recur. What precisely they will discover when they do, and whether the standard models will adapt to the data or whether the data will force something more fundamental to give, is the question that will require years of papers to answer.
