Smartest Bird in the World: Ranked by What Studies Measured

Which is the smartest bird in the world? No bird has an IQ, so the answer depends on the measure. By forebrain neurons in a 2016 count of 28 bird species, the blue-and-yellow macaw leads with 2.46 billion, then the kea, the sulphur-crested cockatoo and the raven at 1.36 billion. On behaviour tests, ravens, crows, rooks, scrub-jays and grey parrots all have standout results. No study we found ranks them on one test.

The new finding is a caveat that rarely makes the headline. The famous neuron count has no crow in it. The New Caledonian crow, star of the tool-use studies, was not among the 28 species, and the top parrots are one bird each. Below: the ranking by neurons, the ranking by tests, and why they do not add up to one winner.

Which is the smartest bird in the world?

No one can say who the smartest bird in the world is. Intelligence has no single scale across species, and no study compares all the contenders. Two kinds of evidence exist. One is how many neurons a bird's forebrain holds. The other is how birds do on tests built for them. Both point to two groups of smartest bird species, corvids (crows, ravens, jays) and parrots, but not to one winner.

For single birds and single skills, our guides to crow and raven IQ and parrot IQ go into detail. This article is the side-by-side view they do not give.

Which birds have the most neurons? The smartest birds ranked by the count

Olkowicz and colleagues (Proceedings of the National Academy of Sciences, 2016) counted neurons in 73 birds from 28 species. The forebrain, or telencephalon, is the part that does the work of the mammalian cortex. Here are the smartest birds ranked by that count, the top ten species and two comparison birds, with the number of birds counted in brackets. Brain mass is in grams and neurons are in millions.

• 1. Blue-and-yellow macaw (1 bird): forebrain 2,459, whole brain 3,136, brain mass 20.7 g
• 2. Kea (1 bird): forebrain 1,630, whole brain 2,149, brain mass 13.6 g
• 3. Sulphur-crested cockatoo (1 bird): forebrain 1,491, whole brain 2,122, brain mass 10.1 g
• 4. Common raven (3 birds): forebrain 1,355, whole brain 2,171, brain mass 14.1 g
• 5. Grey parrot (2 birds): forebrain 1,148, whole brain 1,566, brain mass 8.8 g
• 6. Rook (3 birds): forebrain 918, whole brain 1,509, brain mass 8.4 g
• 7. Goffin's cockatoo (2 birds): forebrain 715, whole brain 1,161, brain mass 8.3 g
• 8. Eurasian jay (3 birds): forebrain 600, whole brain 1,085, brain mass 4.6 g
• 9. Jackdaw (3 birds): forebrain 541, whole brain 968, brain mass 6.0 g
• 10. Magpie (3 birds): forebrain 498, whole brain 897, brain mass 5.4 g
• For comparison, rock pigeon (3 birds): forebrain 83, whole brain 310, brain mass 2.1 g. Red junglefowl (3 birds), the wild ancestor of the chicken: forebrain 74, whole brain 221, brain mass 2.8 g

Three things stand out. The top four species hold 1.4 to 2.5 billion forebrain neurons, against 83 million in a pigeon. A raven's forebrain has about 16 times the pigeon's, and a macaw's about 30 times. Second, the top ten splits five and five between parrots and corvids, and parrots hold four of the top five places. Third, brain mass is a poor guide: the kea, at 13.6 g, has more forebrain neurons than the raven at 14.1 g.

For scale, the paper puts the pallium, the cortex-like layer of the forebrain, at about 1.2 billion neurons in the raven and kea and about 1.9 billion in the macaw. A human cerebral cortex holds about 16 billion, by our arithmetic from the 86.1 billion neurons and 19% cortical share counted by Azevedo and colleagues (2009). The macaw's pallium is about one-eighth of that.

How reliable is the bird neuron count?

Less than the headlines suggest. The macaw, kea and sulphur-crested cockatoo are single birds. The grey parrot and Goffin's cockatoo are two each. The paper's own methods say the split between pallium and subpallium was made in one individual per species. A count from one bird cannot say how much birds of that species vary.

The sample also leaves out the famous names. The 13 songbird species include a raven, a rook, a jackdaw, a magpie, an azure-winged magpie and a Eurasian jay, but no crow. The New Caledonian crow has no neuron count in this paper. Any claim that the crow's tool use "matches its neuron count" is therefore an inference from its cousins.

Which birds do best on cognition tests?

Several, and no bird wins them all. Each result below comes from a different task and a small sample.

• New Caledonian crows: 8 crows treated a moving stick differently when a hidden human might have moved it, which suggests reasoning about an unseen cause (Taylor, Miller and Gray, PNAS, 2012). Earlier work reported crows shaping hooks (Weir, Chappell and Kacelnik, Science, 2002)
• Rooks: all 4 captive rooks solved an Aesop's-fable problem by dropping stones to raise water until a floating worm came within reach (Bird and Emery, Current Biology, 2009)
• Western scrub-jays: the jays stored food where they would be hungry the next morning, which the authors read as planning for a future need (Raby and colleagues, Nature, 2007)
• Magpies: some birds showed mark-directed behaviour in a mirror test, the first such evidence in a non-mammal (Prior, Schwarz and Gunturkun, PLoS Biology, 2008)
• Crows: they produced one to four calls on cue, and the first call predicted the total, which the authors read as planning (Liao, Brecht, Veit and Nieder, Science, 2024)
• Grey parrot: one bird, Alex, named quantities up to six with accuracy above 80% and used "none" for zero (Pepperberg and Gordon, Journal of Comparative Psychology, 2005)
• Kea: given a string-pulling problem, the birds found an immediate solution that nine more trials did not improve, which the review reads as insight (Huber and Gajdon, Animal Cognition, 2006)
• Goffin's cockatoo: one bird spontaneously made and used tools (Auersperg and colleagues, Current Biology, 2012)

Notice the sample sizes: 8 crows, 4 rooks, one grey parrot, one cockatoo. These are strong findings about those birds on those tasks. They are not a league table. For a comparison with dogs, where a ranking does exist for one measure, see our article on smartest dog breeds and what Coren's order measures.

Are crows smarter than parrots?

The evidence does not decide it. Parrots lead on raw forebrain neurons, with the macaw at 2.46 billion against the raven's 1.36 billion. Corvids show strong results on tool use, planning and reasoning about hidden causes. We found no study that tested ravens, grey parrots and keas on the same battery, so a head-to-head claim goes beyond the data.

What the count does show is that both groups sit far above ordinary birds. The paper's abstract says corvids and parrots have much higher proportions of brain neurons in the pallium than primates, other mammals or other birds, and that large-brained parrots and corvids have forebrain counts equal to or greater than primates with much larger brains. That makes them the sensible short list, not the proven winners.

Why can't birds be ranked by IQ?

An IQ is a position on a bell curve built from thousands of people, with an average of 100 and a standard deviation of 15 on most modern tests. No such norm group exists for any bird. The IQ bell curve shows what the number means for people, and why a bird cannot have one.

Researchers compare animals task by task. Our smartest animals study sets out how we handled that across species, and our guide to brain size and IQ explains why brain mass alone is a weak guide even in people. Pigeons, the low end of the neuron table, still do surprising things; see the Monty Hall problem and the pigeons.

What is the smartest bird ever?

No one can name the smartest bird ever. The best-documented individual in the studies above is Alex, the grey parrot, whose number skills were tested in the studies above. But he is one bird tested on one family of tasks, and no study sets him beside the best crow or kea. The smartest bird on earth, as a title, rests on a scale nobody has built.

If you want a number for yourself, take the IQ test and read it against the bell curve. Birds have to be compared task by task. People can be placed on one scale.

Common questions