Astrophysics  ·  4 October 2026

Escape Velocity Calculator: From Leaving Earth to Falling Into a Black Hole

Fire a cannonball fast enough off a mountaintop and it doesn't just fly farther — past a certain speed, it stops coming down at all. Isaac Newton drew exactly this scene in A Treatise of the System of the World to explain orbits, and the same simple idea, pushed a little further, explains something Newton never could have imagined: black holes. Ask the same question — "how fast do you have to go to leave, for good?" — of an object so dense that even light can't go fast enough, and you've arrived at the Schwarzschild radius, the boundary of a black hole.

This page is an escape velocity calculator you fire rather than read. Pick a body — the Moon, Mars, Earth, Jupiter or the Sun — dial in a launch speed and fire. The ball crashes back down, loops forever without landing, or leaves and never returns, and every shot's trail stays on screen until you have rebuilt Newton's own diagram out of your own attempts. Below that sits a press. It crushes a body at constant mass while you watch what can still get off a shrinking surface, each speed class failing in turn until even light can't manage it. When light gives up, you've made a black hole, and the radius where that happens is the Schwarzschild radius.

What Is Escape Velocity and How Do You Calculate It?

Escape velocity is the minimum speed an object needs — launched once, with no further thrust — to break free of a body's gravity forever. Give it exactly that speed and it just barely makes it to infinity, arriving there with essentially nothing left. Give it more, and it leaves with speed to spare.

The formula falls out of a straightforward energy argument. An object of mass mm has kinetic energy 12mv2\frac{1}{2}mv^2 and gravitational potential energy −GMmr-\frac{GMm}{r} (negative, because gravity is attractive — it takes work to pull the object away, not push it). "Just barely escaping" means the total mechanical energy is exactly zero: kinetic energy exactly cancels the gravitational binding energy at the surface.

12mvesc2=GMmR\frac{1}{2}mv_{esc}^2 = \frac{GMm}{R}

The object's own mass mm cancels out completely. Solve for vv and you get:

vesc=2GMRv_{esc} = \sqrt{\frac{2GM}{R}}

where MM is the mass of the body you're escaping, RR its radius, and GG the gravitational constant, 6.674×10−11 N⋅m2/kg26.674 \times 10^{-11}\ \text{N·m}^2/\text{kg}^2. Notice what's missing from that formula: the mass of whatever's escaping. A feather and a rocket need exactly the same escape velocity — the difference is entirely in how much energy it takes to get each of them up to that speed, not in the speed itself.

BodyEscape velocityOrbital velocity (low orbit)
Moon2.38 km/s1.68 km/s
Mars5.03 km/s3.55 km/s
Earth11.19 km/s7.91 km/s
Jupiter59.53 km/s42.10 km/s
Sun617.75 km/s436.82 km/s

Earth's escape velocity, 11.19 km/s, is worth sitting with for a second — that's just over 40,000 km/h, or about 33 times the speed of sound. No rocket actually needs to hit that speed while still down at the surface (they build up speed gradually, fighting less gravity the higher they climb), but 11.19 km/s is what a single unpowered shot would need.

Does Escape Velocity Depend on Direction or Mass?

Not on the mass of the escaping object, as the derivation above shows. Direction matters less than you'd think, too: fired straight up, an object needs exactly vescv_{esc}. Fired at an angle, or even sideways, it still escapes at exactly the same speed — what changes is the shape of the path getting there, not the speed threshold itself. That's a direct consequence of energy conservation: "just enough energy to reach infinity" doesn't care which way the kinetic energy happens to point.

Newton's Cannon: Why Speed Alone Decides Orbit or Escape

Here's the thought experiment Newton actually used to argue that orbits and falling apples are the same physics. Imagine a cannon on a mountain so tall its tip pokes above the atmosphere, firing cannonballs perfectly horizontally.

Fire slowly, and the cannonball arcs out a little and lands — ordinary projectile motion, gravity pulling it down while it drifts forward. Fire faster and it goes farther before landing, the arc flattening out. Keep increasing the speed, and at some point the ball is falling around the curve of the planet exactly as fast as the ground curves away beneath it — it never lands. That's an orbit. Push the speed higher still, past 2\sqrt{2} times that orbital speed, and gravity can no longer bend the path into a closed loop at all: the cannonball leaves and keeps leaving, forever.

Newton could only draw that figure. Below, you get to fire it — one shot at a time. The strip under the picture names both boundaries from the start; what it cannot tell you is which side of them any particular shot falls on, and that is what firing is for.

Try it yourself

  1. It opens on Earth with the launch-speed dial at 0.50 × v_esc and an empty sky. Press Fire. The ball leaves the mountaintop sideways at 5.46 km/s and starts falling immediately — a horizontal shot this slow is already at the top of its ellipse, so it never gains a metre of height. It sweeps about 18° around the world on the way down, and the whole trail turns red the moment it hits. Read what the banner says about that shot: it was on an orbit the entire flight. The ground simply arrived before the far end of the ellipse did.
  2. Put the dial on 0.75 × and fire again. Nothing has changed but the number — 8.19 km/s instead of 5.46 — and this time the ball falls all the way round and misses. The trail closes into an amber loop that never touches the ground. That is an orbit: the same fall as before, with enough sideways speed that the surface keeps curving out from under it.
  3. Now 1.05 ×, which is 11.46 km/s. The trail goes green, runs off the edge of the picture with an arrowhead on it, and that is the last you see of it. Gravity never stops pulling on that ball. It just never manages to turn it round.
  4. Go hunting for the two boundaries. Fire a spread — 0.60, 0.68, 0.72, 0.90, 0.93, 1.00 — then stop watching the sky and look at the strip along the bottom edge instead. Every resolved shot has left a tick there at the speed it was fired at, red, amber or green, and your own ticks close in on the two dashed lines from both sides: v_esc/√2, about 0.71, where crashing turns into orbiting, and 1.00, where orbiting turns into leaving. The real crash-to-orbit boundary sits a shade below the 1/√2 line — at 0.698 rather than 0.707 — because the cannon stands 5% above the surface, so a shot just under circular speed can still skim past the ground on the far side. Fire 0.70 to see it: a shade under 1/√2, and it orbits.
  5. Try 0.98 ×. The ball climbs until the body is a dot, runs clean off the picture, and the banner still reads STILL BOUND — with a hollow arrowhead where an escapee would have a solid one. It has less kinetic energy than the well has depth, so however far it gets, it is coming back. Going a very long way is not the same as getting away, and that is what makes the threshold sharp.
  6. Switch the Body dropdown to the Moon. The volley clears — every trail is drawn in units of the body it was fired from — and the whole story replays at roughly a fifth of the speed: 1.16 km/s crashes, 1.74 km/s orbits, 2.43 km/s leaves for good. The fractions are identical because the fractions are the physics. Only the Moon's own numbers changed.
Set a speed and fire the cannon.
Each shot leaves the mountaintop horizontally at your chosen fraction of escape velocity. Bracket the two dashed lines on the strip below and you will find both thresholds yourself.
Earth at r₀ = 1.05 R — v_esc = 10.92 km/s · v_orbit = 7.72 km/s
The cannon stands on Newton’s “very high mountain” at r₀ = 1.05 R, not on the surface, so every speed quoted here is the value at that altitude — about 2.4% below the textbook surface figure (11.19 km/s for Earth). That is what puts 1.00 × exactly on the escape boundary in this picture. Switching bodies clears the volley, because every trail is drawn in units of the body it was fired from. A path is also only followed so far — out to about twenty body radii, or until a fixed integration budget runs out, whichever ends first — and that is a limit on the drawing, never on the verdict: from just below 0.97 × upward the drawing stops while the ball is still out there, and the shot is nonetheless bound. You will see it end in a hollow arrowhead and read as STILL BOUND, because what settles a shot’s fate is whether its kinetic energy covers the depth of the well, not how far it happened to get before the picture ran out. The dividing line is exactly 1.00 ×. Two-body Newtonian gravity only: no atmosphere, no rotating ground, no other worlds pulling — and the ball’s own mass never appears anywhere, which is why a cannonball and a spacecraft need exactly the same speed.
Body
Switching bodies clears the volley — the picture is drawn in units of this body’s radius.
Launch speed
0.50 × v_esc
5.46 km/s — fired horizontally from the summit, with no engine and nothing else pushing it.
Fire again at any time — a shot still in the air resolves at once and the next one leaves immediately. Every shot plays at its own steady pace, so the ball really does speed up near the body and crawl at the far end of its arc.
body radius R6,371 km
launch altitude r₀ = 1.05 R6,690 km
launch speed5.46 km/s
v_orbit(r₀) = v_esc/√27.72 km/s
v_esc(r₀)10.92 km/s
shots — crash / orbit / escape0 / 0 / 0

Fire a dozen shots and you will have redrawn, by hand, the figure Newton put in his System of the World: a fan of arcs from one mountaintop, each reaching further round the world than the last, until one of them stops coming down. The tick strip underneath is the same discovery written as data: red gives way to amber at one dashed line, amber to green at the other.

One honest detail about the numbers. The cannon stands on Newton's mountain rather than on the ground: r0=1.05Rr_0 = 1.05R, five per cent up. That isn't decoration — a horizontal shot fired from exactly RR at less than orbital speed starts at the top of its ellipse and dips below the surface immediately, so every slow shot would bury itself before you saw an arc. Every speed the widget quotes is therefore the value at that altitude, about 2.4% below the textbook surface figure in the table above: 10.92 km/s for Earth rather than 11.19. The physics is identical. And because every speed on the dial is measured against v_esc at that same launch altitude, the 1.00 × mark sits exactly on the escape boundary you can watch happen — the altitude moves the number, not the threshold.

What Is the Difference Between Orbital Velocity and Escape Velocity?

Both numbers come from the exact same balance of forces, just answering a different question. Orbital velocity asks "how fast to fall in a circle around this radius?" — set gravitational force equal to the centripetal force requirement (see our centripetal force calculator for that derivation on its own) and you get vorbit=GM/Rv_{orbit} = \sqrt{GM/R}. Escape velocity asks a bigger question — "how fast to leave the gravity well altogether?" — and the answer is always exactly 2\sqrt{2} times larger:

vesc=2 vorbitv_{esc} = \sqrt{2}\, v_{orbit}

That factor of 2≈1.414\sqrt{2} \approx 1.414 isn't a coincidence or a rough approximation — it drops straight out of comparing the two energy equations, and it's exactly why the amber ticks on the strip above give way to green where they do. It has a bit of history behind it, too: Soviet engineers labelled these the "cosmic velocities" — orbital speed is the first cosmic velocity, escape speed the second. For a full treatment of what a stable orbit actually looks like over time — ellipses, periods, Kepler's laws — see our orbital mechanics post; this page is deliberately about the boundary case, not the stable orbits sitting on either side of it.

How Do Black Holes Form?

A star spends most of its life in a standoff: gravity crushing inward, the outward pressure from nuclear fusion pushing back. Fusion wins that argument for billions of years — until the fuel runs out. For an ordinary star like the Sun, gravity wins gently: the core settles into a white dwarf, held up by quantum pressure between crowded electrons, and that's the end of the story.

For a star roughly 20 times the Sun's mass or more, gravity wins violently. The core collapses in on itself in under a second, the outer layers detonate outward as a supernova, and whatever's left of the core keeps falling inward with nothing left to stop it. If that remaining core is heavier than roughly 3 solar masses, not even the exotic pressure that supports a neutron star can hold the line. The collapse doesn't stop. A black hole is what "doesn't stop" looks like: matter compressed past the point where its own escape velocity would exceed the speed of light.

That last clause is the entire definition — and it connects straight back to the formula from the first section.

What Is the Schwarzschild Radius?

Take the escape velocity formula and ask a slightly unsettling question: what radius would a given mass MM need to be squeezed into for its escape velocity to equal cc, the speed of light — the fastest anything can ever travel?

c=2GMRsc = \sqrt{\frac{2GM}{R_s}}

Square both sides and solve for RR:

Rs=2GMc2R_s = \frac{2GM}{c^2}

This is the Schwarzschild radius, named for Karl Schwarzschild, who derived it properly using Einstein's general relativity in 1916 — remarkably, working out an exact solution to Einstein's field equations for a spherical mass while serving on the Russian front in the First World War, only months before his death. The Newtonian derivation above isn't technically rigorous (photons don't have mass, and gravity this close to a black hole isn't Newtonian at all), but it lands on precisely the right answer anyway — one of those odd coincidences in physics that turns out not to be a coincidence once you understand the deeper theory underneath it.

Compress any mass down inside its own Schwarzschild radius and the result is a black hole: a region where the escape velocity genuinely exceeds the speed of light, so nothing — not a rocket, not a radio signal, not light itself — can climb back out. The boundary itself, at radius RsR_s, is called the event horizon.

That is easy enough to write down and very hard to feel. The press below is an attempt at the feeling: one slider squeezing one body at a mass that never changes, its surface firing test launches the entire time you crush it.

Try it yourself

  1. It opens on Earth at its true size, R = 6,371 km, with the compression slider at zero. The surface is spraying short streaks in every direction at five fixed real speeds: a chemical rocket at 11.19 km/s, the escape speed at the Sun's own surface, 1% of light speed, 10% of it, and light itself. At this size all five classes get away.
  2. Nudge the slider one notch. That is all it takes — the rocket streaks stop leaving. They climb, slow, stop and fall back, because a slightly smaller Earth with exactly the same mass inside has an escape velocity slightly above 11.19 km/s, and the rocket class was only ever precisely at it.
  3. Keep squeezing and watch the legend go dark from the top down. The solar class, fast enough to leave the surface of the Sun, dies at about 3,000× compression. The 1%-of-light class survives to roughly 72,000×, and the 10% class to about 7 million×. Nothing here is being aimed: the spray is radially symmetric on purpose, because escape velocity does not care which way you point.
  4. Push the slider to the far end of its travel, just past 8.85. The light streaks curl back too, the disc snaps black, the dashed ring lights up, and the banner reports that Earth crossed its horizon at 7.2 × 10⁸× compression. That ring never moved through any of this — Rₛ = 2GM/c² depends on the mass alone, and you never touched the mass.
  5. Now the part worth the whole widget. Switch the body to Neutron star, then drag the compression slider back down to 0 first — switching bodies does not reset it, so left where it was you would land on a surface already crushed far inside its own horizon. At its true size — 1.4 solar masses inside a 12 km radius — the surface's own escape velocity is already about 176,000 km/s, 59% of the speed of light: the rocket, solar, 1%-c and 10%-c classes are already falling back, and only the light class is still getting out. Squeeze from there and it collapses after barely a nudge — a factor of 2.9 finishes it, horizon at just 0.46 of the same 9.5-unit travel Earth needed nearly all of. Now park the slider anywhere and flip between the three bodies: at one fixed setting you can see instantly which of them is already a black hole and which is nowhere close.
Earth at its true size — R = 6,371 km.
Nothing has been squeezed yet. v_esc at the surface is 11.19 km/s, which is 0.0037% of c. All five launch speeds get away. Now start crushing it: the mass stays exactly where it is, only R shrinks.
Mean density ρ = 5,513 kg/m³. Denser than water — still ordinary matter, held apart by ordinary electron shells.
The radial scale is not distance and could not be: an escaping streak goes to infinity. It is the fraction of the gravity well climbed, 1 − R/r, so the outer dashed ring is r = ∞ and a streak launched at a fraction u of escape velocity tops out at exactly u² of the way there — which is why a class at 70% of v_esc stops just under halfway up rather than 70% of the way. The body’s own drawn radius is logarithmic between its true size and the horizon ring, because nine orders of magnitude of shrinking will not fit on a linear one; the horizon ring itself never moves, since Rₛ = 2GM/c² depends on the mass alone.Each streak’s motion is the exact radial two-body solution, so the way it slows is real — but the wall-clock pace is compressed class by class, and has to be. The five launch speeds span a factor of 26,800, and at this compression their true flight times span a factor of 56,200; the picture spans a factor of 4.3. Paced honestly against each other they would be four invisible flickers and one streak that never appeared to move. What is exact is the verdict: a streak leaves if and only if its speed reaches v_esc at the current surface, whatever direction it points. The press is Newtonian throughout, and deliberately so — this is Michell’s 1783 “dark star”, which Laplace reached independently in 1796: light treated as a projectile that can be slowed. Real light never slows; general relativity gets the same Rₛ = 2GM/c² by an entirely different route, and inside a horizon the Newtonian picture stops meaning anything at all. Density assumes a uniform sphere, and real matter fights back long before these numbers: the milestones say where, and no known force resists past the last of them.
Body in the press
Compression
0 decades
1.0× — radius now 6,371 km, escape velocity 11.19 km/s.
mass M (never changes)1.00 Earth masses
true radius R₀6,371 km
current radius R6,371 km
compression R₀/R1.0×
Rₛ = 2GM/c²8.87 mm
R₀ / Rₛ (squeeze to collapse)7.2 × 10⁸×
v_esc at R11.19 km/s
mean density ρ5,513 kg/m³

Two things in that sequence are worth pulling apart. That the classes fail in speed order is obvious the moment you see it. What isn't obvious is that the last class to fail is special: light's threshold isn't one milestone among five, it is the definition. Setting vesc=cv_{esc} = c is the equation Rs=2GM/c2R_s = 2GM/c^2 solves, so when the light streaks curl back the surface has reached the ring, and nothing below it will ever be heard from outside again. (The press is Newtonian throughout, which is John Michell's "dark star" of 1783 — an idea Laplace arrived at independently in 1796 — rather than Einstein's black hole. Real light doesn't slow down. General relativity arrives at the same RsR_s by an entirely different route, which is the coincidence-that-isn't from a moment ago.)

How Small Would Earth Have to Be to Become a Black Hole?

BodyActual radiusSchwarzschild radiusCompression needed
Moon1,737 km0.109 mm~1.6 × 10¹⁰×
Mars3,390 km0.953 mm~3.6 × 10⁹×
Earth6,371 km8.87 mm~7.2 × 10⁸×
Jupiter71,492 km2.82 m~2.5 × 10⁷×
Sun695,700 km2.95 km~2.4 × 10⁵×

Earth's entire mass, squeezed into a sphere just 8.87 mm in radius — about the size of a marble — would become a black hole. The Sun would need to shrink to a radius of about 2.95 km: a sphere 5.9 km across, a few minutes' drive at highway speed, still holding onto its full mass.

Now put a neutron star next to that column. It packs about 1.4 solar masses — 2.78×10302.78 \times 10^{30} kg — into a ball roughly 12 km in radius, and its Schwarzschild radius is 4.14 km. The compression it still needs is a factor of 2.9. Earth is nearly nine orders of magnitude from its own horizon; a dead stellar core has covered all but the last factor of three of that same distance. That gap, not the mass, is the whole reason black holes come from collapsed stars and never from planets. Nothing in the ordinary universe squeezes a planet at all, while a star's own collapse routinely does almost the entire job — and dropping a little more mass onto a neutron star finishes it without anyone's help.

Real Black Holes: From Stellar Remnants to Sagittarius A*

The smallest black holes that can form from stellar collapse sit right around 3 solar masses — the theoretical floor below which the leftover pressure in a collapsing core can still hold the line as a neutron star instead. A black hole that size has a Schwarzschild radius of about 8.9 km — an event horizon roughly 18 km across, about the span of a city.

Far larger black holes sit at the centers of galaxies. Sagittarius A*, the supermassive black hole anchoring the Milky Way, weighs in at about 4.3 million solar masses — its Schwarzschild radius works out to roughly 12.7 million km, comfortably smaller than Mercury's 0.39 AU orbit around the Sun. M87*, the first black hole ever directly imaged — by the Event Horizon Telescope in 2019 — is far bigger still: about 6.5 billion solar masses, giving it a Schwarzschild radius of roughly 128 AU, bigger than Neptune's entire 30 AU orbit. Swap M87* in for our Sun and its event horizon alone would swallow every planet in the solar system. Working out distances on this scale — to the galactic center, to a galaxy 55 million light-years away — leans on the same cosmic distance ladder covered in our stellar distances post, just stretched to its most extreme rungs.

Worked Examples for Physics Exams

Worked Example

Example 1 — Escape velocity from Earth's surface

What speed must a projectile reach at Earth's surface, with no further thrust, to escape Earth's gravity entirely?

vesc=2GMR=2×6.674×10−11×5.972×10246.371×106≈11,190 m/s=11.19 km/sv_{esc} = \sqrt{\frac{2GM}{R}} = \sqrt{\frac{2 \times 6.674\times10^{-11} \times 5.972\times10^{24}}{6.371\times10^{6}}} \approx 11{,}190\ \text{m/s} = 11.19\ \text{km/s}

That's about 40,270 km/h — roughly 33 times the speed of sound at sea level.

Worked Example

Example 2 — Why the Apollo Lunar Module needed so little fuel to leave the Moon

The Apollo Lunar Module's ascent stage, launching astronauts off the Moon back to lunar orbit, used a strikingly small engine compared to the Saturn V that launched from Earth. Why?

The Moon's escape velocity is 2×6.674×10−11×7.342×1022/1.737×106≈2,375\sqrt{2 \times 6.674\times10^{-11} \times 7.342\times10^{22} / 1.737\times10^{6}} \approx 2{,}375 m/s =2.38= 2.38 km/s — against Earth's 11.19 km/s, a factor of about 4.7 times less. Since the energy required scales with the square of the speed, leaving the Moon takes roughly 4.72≈224.7^2 \approx 22 times less energy per kilogram than leaving Earth. That's the whole reason the ascent stage could be not much bigger than a garden shed — about 4.3 m across and 2.8 m tall — bolted to a descent stage, while getting off Earth took a rocket like the Saturn V, itself about 111 meters tall.

Worked Example

Example 3 — Deriving the Sun's Schwarzschild radius

If the Sun's entire mass were compressed until its escape velocity reached the speed of light, what radius would it have?

Rs=2GMc2=2×6.674×10−11×1.989×1030(2.998×108)2≈2,954 m≈2.95 kmR_s = \frac{2GM}{c^2} = \frac{2 \times 6.674\times10^{-11} \times 1.989\times10^{30}}{(2.998\times10^8)^2} \approx 2{,}954\ \text{m} \approx 2.95\ \text{km}

Compare that to the Sun's actual radius of 695,700 km, and you get a sense of just how extreme "black hole density" really is: about 236,000 times smaller across, with every bit of the same mass packed inside.

Frequently Asked Questions

Explore more simulations

Every concept on PhysicStuff has an interactive simulation. No login, no setup required.

See how every topic connects on the concept map.