Best cities in the world to visit

We asked 10 models: The best cities in the world to visit as a tourist. Each named its own top 10.

Agreement between models

67%

Same model, run twice

80%

Different picks

16

When agreement between models is well below a model's agreement with itself, the split reflects real differences in taste rather than randomness.

The consensus

Each cell is the position that model gave the pick; blank means it left the pick off. Click a row for each model's reasoning.

#	Pick	Score	Listed by	Claude Haiku 5.5	GPT-6 Luna	Gemini 4 Argon	Qwen 3.8 Flash	DeepSeek V4 Pro	GLM-5.3	Kimi K3	MiniMax M3	Mistral Large 4	Nemotron 3 Ultra
1	Tokyo		10/10	3	1	4	4	1	2	1	2	2	2
2	Paris		10/10	1	2	3	3	3	4	2	1	5	1
3	Rome		10/10	2	3	1	2	2	1	3	3	4	4
4	Kyoto		10/10	4	6	2	1	6	6	7	6	1	7
5	Barcelona		10/10	5	7	8	9	4	5	5	5	3	5
6	Istanbul		10/10	6	4	5	5	5	3	6	7	9	6
7	New York City		10/10	7	5	6	7	7	7	4	4	8	3
8	London		7/10	8	9	7				8	8	10	8
9	Cape Town		6/10		10		8	8	10		10	6
10	Lisbon		5/10	9			10		8		9	7
11	Prague		2/10				6			10
12	Mexico City		3/10		8	9		10
13	Bangkok		3/10			10		9					9
14	Cairo		1/10						9
15	Marrakesh		1/10							9

1 lower-scoring picks are not shown; every ballot below lists them in full.

Every ballot

Hover a pick to find it on the other ballots. Marked picks appear on no other ballot.

Claude Haiku 5.5

1. Paris
2. Rome
3. Tokyo
4. Kyoto
5. Barcelona
6. Istanbul
7. New York City
8. London
9. Lisbon
10. Sydney

76% with consensus · 93% self-consistent

GPT-6 Luna

1. Tokyo
2. Paris
3. Rome
4. Istanbul
5. New York City
6. Kyoto
7. Barcelona
8. Mexico City
9. London
10. Cape Town

89% with consensus · 94% self-consistent

Gemini 4 Argon

1. Rome
2. Kyoto
3. Paris
4. Tokyo
5. Istanbul
6. New York City
7. London
8. Barcelona
9. Mexico City
10. Bangkok

60% with consensus · 74% self-consistent

Qwen 3.8 Flash

1. Kyoto
2. Rome
3. Paris
4. Tokyo
5. Istanbul
6. Prague
7. New York City
8. Cape Town
9. Barcelona
10. Lisbon

59% with consensus · 60% self-consistent

DeepSeek V4 Pro

1. Tokyo
2. Rome
3. Paris
4. Barcelona
5. Istanbul
6. Kyoto
7. New York City
8. Cape Town
9. Bangkok
10. Mexico City

85% with consensus · 80% self-consistent

GLM-5.3

1. Rome
2. Tokyo
3. Istanbul
4. Paris
5. Barcelona
6. Kyoto
7. New York City
8. Lisbon
9. Cairo
10. Cape Town

66% with consensus · 77% self-consistent

Kimi K3

1. Tokyo
2. Paris
3. Rome
4. New York City
5. Barcelona
6. Istanbul
7. Kyoto
8. London
9. Marrakesh
10. Prague

92% with consensus · 93% self-consistent

MiniMax M3

1. Paris
2. Tokyo
3. Rome
4. New York City
5. Barcelona
6. Kyoto
7. Istanbul
8. London
9. Lisbon
10. Cape Town

78% with consensus · 86% self-consistent

Mistral Large 4

1. Kyoto
2. Tokyo
3. Barcelona
4. Rome
5. Paris
6. Cape Town
7. Lisbon
8. New York City
9. Istanbul
10. London

60% with consensus · 61% self-consistent

Nemotron 3 Ultra

1. Paris
2. Tokyo
3. New York City
4. Rome
5. Barcelona
6. Istanbul
7. Kyoto
8. London
9. Bangkok
10. Sydney

72% with consensus · 80% self-consistent

Who agrees with whom

Overlap between two models' lists, weighted toward the top · 100 means identical · Color runs from 46 (lightest) to 92 (darkest)

	Claude Haiku 5.5	GPT-6 Luna	Gemini 4 Argon	Qwen 3.8 Flash	DeepSeek V4 Pro	GLM-5.3	Kimi K3	MiniMax M3	Mistral Large 4	Nemotron 3 Ultra
Claude Haiku 5.5		67	67	64	69	66	70	86	53	81
GPT-6 Luna	67		59	55	84	67	91	76	51	71
Gemini 4 Argon	67	59		77	63	73	55	55	54	52
Qwen 3.8 Flash	64	55	77		63	58	51	52	70	46
DeepSeek V4 Pro	69	84	63	63		74	83	68	58	64
GLM-5.3	66	67	73	58	74		63	65	56	59
Kimi K3	70	91	55	51	83	63		81	52	79
MiniMax M3	86	76	55	52	68	65	81		55	92
Mistral Large 4	53	51	54	70	58	56	52	55		52
Nemotron 3 Ultra	81	71	52	46	64	59	79	92	52