Best first programming language

We asked 10 models: Rank these programming languages from best to worst first language for a complete beginner. Each ranked the same fixed set of options.

Agreement between models

87%

Same model, run twice

92%

Different picks

12

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	Python		10/10	1	2	1	2	2	1	1	1	2	1
2	Scratch		10/10	2	1	3	1	1	2	2	2	1	2
3	JavaScript		10/10	4	3	2	4	5	3	3	3	5	3
4	Ruby		10/10	3	4	4	3	3	4	4	4	3	4
5	Lua		10/10	5	5	5	5	4	5	5	5	4	6
6	Go		10/10	6	6	7	6	6	8	8	6	8	5
7	C#		10/10	8	7	6	7	7	6	6	7	6	8
8	Java		10/10	7	8	8	8	8	7	7	8	7	7
9	C		10/10	10	9	9	10	9	9	9	9	9	9
10	C++		10/10	11	11	12	9	10	10	10	11	11	10
11	Haskell		10/10	9	10	10	12	12	11	12	10	12	12
12	Rust		10/10	12	12	11	11	11	12	11	12	10	11

Every ballot

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

Claude Haiku 5.5

1. Python
2. Scratch
3. Ruby
4. JavaScript
5. Lua
6. Go
7. Java
8. C#
9. Haskell
10. C
11. C++
12. Rust

94% with consensus · 98% self-consistent

GPT-6 Luna

1. Scratch
2. Python
3. JavaScript
4. Ruby
5. Lua
6. Go
7. C#
8. Java
9. C
10. Haskell
11. C++
12. Rust

86% with consensus · 89% self-consistent

Gemini 4 Argon

1. Python
2. JavaScript
3. Scratch
4. Ruby
5. Lua
6. C#
7. Go
8. Java
9. C
10. Haskell
11. Rust
12. C++

91% with consensus · 94% self-consistent

Qwen 3.8 Flash

1. Scratch
2. Python
3. Ruby
4. JavaScript
5. Lua
6. Go
7. C#
8. Java
9. C++
10. C
11. Rust
12. Haskell

81% with consensus · 87% self-consistent

DeepSeek V4 Pro

1. Scratch
2. Python
3. Ruby
4. Lua
5. JavaScript
6. Go
7. C#
8. Java
9. C
10. C++
11. Rust
12. Haskell

79% with consensus · 97% self-consistent

GLM-5.3

1. Python
2. Scratch
3. JavaScript
4. Ruby
5. Lua
6. C#
7. Java
8. Go
9. C
10. C++
11. Haskell
12. Rust

98% with consensus · 97% self-consistent

Kimi K3

1. Python
2. Scratch
3. JavaScript
4. Ruby
5. Lua
6. C#
7. Java
8. Go
9. C
10. C++
11. Rust
12. Haskell

97% with consensus · 99% self-consistent

MiniMax M3

1. Python
2. Scratch
3. JavaScript
4. Ruby
5. Lua
6. Go
7. C#
8. Java
9. C
10. Haskell
11. C++
12. Rust

100% with consensus · 89% self-consistent

Mistral Large 4

1. Scratch
2. Python
3. Ruby
4. Lua
5. JavaScript
6. C#
7. Java
8. Go
9. C
10. Rust
11. C++
12. Haskell

76% with consensus · 86% self-consistent

Nemotron 3 Ultra

1. Python
2. Scratch
3. JavaScript
4. Ruby
5. Go
6. Lua
7. Java
8. C#
9. C
10. C++
11. Rust
12. Haskell

97% with consensus · 84% self-consistent

Who agrees with whom

Overlap between two models' lists, weighted toward the top · 100 means identical · Color runs from 71 (lightest) to 100 (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		81	86	83	81	93	92	95	79	93
GPT-6 Luna	81		78	95	93	83	83	86	90	82
Gemini 4 Argon	86	78		73	71	92	92	92	72	89
Qwen 3.8 Flash	83	95	73		97	79	79	81	94	79
DeepSeek V4 Pro	81	93	71	97		77	77	79	97	77
GLM-5.3	93	83	92	79	77		100	97	79	95
Kimi K3	92	83	92	79	77	100		97	79	96
MiniMax M3	95	86	92	81	79	97	97		76	96
Mistral Large 4	79	90	72	94	97	79	79	76		75
Nemotron 3 Ultra	93	82	89	79	77	95	96	96	75