Best national cuisines

We asked 10 models: The best national cuisines in the world. Each named its own top 10.

Agreement between models

76%

Same model, run twice

91%

Different picks

13

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	Italian		10/10	1	1	1	1	1	2	2	3	1	1
2	Japanese		10/10	2	2	3	2	2	3	1	1	2	2
3	Mexican		10/10	6	3	2	6	3	4	3	7	6	3
4	Thai		10/10	4	4	5	4	6	1	6	5	7	5
5	Indian		10/10	7	5	4	5	4	5	5	6	5	4
6	Chinese		10/10	3	6	7	3	7	6	4	4	4	7
7	French		10/10	5	7	6	7	5	7	7	2	3	6
8	Korean		8/10	9			8	10	10	9	9	10	8
9	Spanish		5/10		10	8					8	8	10
10	Turkish		6/10	10	8		9	9	9	10
11	Vietnamese		4/10		9	10			8	8
12	Lebanese		4/10			9		8			10	9
13	Greek		3/10	8			10						9

Every ballot

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

Claude Haiku 5.5

1. Italian
2. Japanese
3. Chinese
4. Thai
5. French
6. Mexican
7. Indian
8. Greek
9. Korean
10. Turkish

85% with consensus · 97% self-consistent

GPT-6 Luna

1. Italian
2. Japanese
3. Mexican
4. Thai
5. Indian
6. Chinese
7. French
8. Turkish
9. Vietnamese
10. Spanish

97% with consensus · 99% self-consistent

Gemini 4 Argon

1. Italian
2. Mexican
3. Japanese
4. Indian
5. Thai
6. French
7. Chinese
8. Spanish
9. Lebanese
10. Vietnamese

86% with consensus · 90% self-consistent

Qwen 3.8 Flash

1. Italian
2. Japanese
3. Chinese
4. Thai
5. Indian
6. Mexican
7. French
8. Korean
9. Turkish
10. Greek

90% with consensus · 93% self-consistent

DeepSeek V4 Pro

1. Italian
2. Japanese
3. Mexican
4. Indian
5. French
6. Thai
7. Chinese
8. Lebanese
9. Turkish
10. Korean

91% with consensus · 92% self-consistent

GLM-5.3

1. Thai
2. Italian
3. Japanese
4. Mexican
5. Indian
6. Chinese
7. French
8. Vietnamese
9. Turkish
10. Korean

71% with consensus · 84% self-consistent

Kimi K3

1. Japanese
2. Italian
3. Mexican
4. Chinese
5. Indian
6. Thai
7. French
8. Vietnamese
9. Korean
10. Turkish

78% with consensus · 80% self-consistent

MiniMax M3

1. Japanese
2. French
3. Italian
4. Chinese
5. Thai
6. Indian
7. Mexican
8. Spanish
9. Korean
10. Lebanese

61% with consensus · 94% self-consistent

Mistral Large 4

1. Italian
2. Japanese
3. French
4. Chinese
5. Indian
6. Mexican
7. Thai
8. Spanish
9. Lebanese
10. Korean

83% with consensus · 80% self-consistent

Nemotron 3 Ultra

1. Italian
2. Japanese
3. Mexican
4. Indian
5. Thai
6. French
7. Chinese
8. Korean
9. Greek
10. Spanish

94% with consensus · 96% self-consistent

Who agrees with whom

Overlap between two models' lists, weighted toward the top · 100 means identical · Color runs from 52 (lightest) to 95 (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		84	74	95	82	62	70	66	86	83
GPT-6 Luna	84		86	88	91	72	78	59	81	92
Gemini 4 Argon	74	86		76	88	66	68	54	77	90
Qwen 3.8 Flash	95	88	76		83	67	73	64	86	85
DeepSeek V4 Pro	82	91	88	83		65	75	60	85	94
GLM-5.3	62	72	66	67	65		68	52	59	66
Kimi K3	70	78	68	73	75	68		78	71	76
MiniMax M3	66	59	54	64	60	52	78		74	60
Mistral Large 4	86	81	77	86	85	59	71	74		82
Nemotron 3 Ultra	83	92	90	85	94	66	76	60	82