Model comparison

Codestral vs Llama 3.2 1B

Codestral is the stronger model overall, scoring 30.6 to 20.1 on the Noometry Index. Llama 3.2 1B costs 6.4× less per token, which makes it the better buy when Codestral's lead doesn't matter for your workload.

Last verified October 10, 2026. 2 shared benchmarks.

Codestral

Mistral AI

30.6

Rank #290 Reported

Llama 3.2 1B

Summary

They share 2 benchmarks with published results for both. Codestral scores higher in 2 categories and Llama 3.2 1B in 0 categories; 2 gaps are clear of the uncertainty.
The widest gap is in coding, where Codestral leads 27.3 to 21.1.
The biggest single-benchmark swing is BigCodeBench Complete: 52.5% for Codestral and 11.3% for Llama 3.2 1B.
Llama 3.2 1B is cheaper at $0.027 / $0.20 per million input/output tokens, against $0.30 / $0.90 for Codestral.
Codestral accepts more context: 256K tokens versus 60K.
Llama 3.2 1B has downloadable open weights; the other is API-only.

Side by side

Codestral and Llama 3.2 1B specifications
	Codestral	Llama 3.2 1B
Provider	Mistral AI	Meta
Noometry Index	30.6	20.1
Released	2024-05-29	2024-09-24
Weights	Proprietary	Open
Context window	256K	60K
Max output	8K	54K
Input $ / M tokens	$0.30	$0.027
Output $ / M tokens	$0.90	$0.20
Results tracked	7	22

Sponsored placements are available on pages like this one. Advertise on Noometry

Category by category

Coding Codestral leads

Codestral: 27.3 (#321), Llama 3.2 1B: 21.1 (#338)

Coding benchmarks
Benchmark	Codestral	Llama 3.2 1B
BigCodeBench Instruct	41.8%	8.2%
BigCodeBench Complete	52.5%	11.3%
Aider Polyglot	11.1%	—
LMArena Coding	—	1070
ALE-Bench	137.78	—
HumanEval+	73.8%	—
MBPP+	61.9%	—

Agentic & Tool Use Not comparable

Codestral: —, Llama 3.2 1B: 14.6 (#150)

Agentic & Tool Use benchmarks
Benchmark	Codestral	Llama 3.2 1B
Berkeley Function Calling Leaderboard	—	10.8%
BALROG	—	6.6%

Reasoning Codestral leads

Codestral: 19.8 (#251), Llama 3.2 1B: 16.2 (#308)

Reasoning benchmarks
Benchmark	Codestral	Llama 3.2 1B
Kagi LLM Benchmark	32.5%	—
Chess Puzzles	—	0%
LMArena Hard Prompts	—	1044
Epoch Capabilities Index	—	101.99

Math Not comparable

Codestral: —, Llama 3.2 1B: 10.4 (#313)

Math benchmarks
Benchmark	Codestral	Llama 3.2 1B
OTIS Mock AIME 2024-2025	—	0.6%
LMArena Math	—	1086

Knowledge Not comparable

Codestral: —, Llama 3.2 1B: 7.2 (#312)

Knowledge benchmarks
Benchmark	Codestral	Llama 3.2 1B
GPQA Diamond	—	23.9%
LMArena Expert	—	1007

Multilingual Not comparable

Codestral: —, Llama 3.2 1B: 23.8 (#292)

Multilingual benchmarks
Benchmark	Codestral	Llama 3.2 1B
LMArena Non-English	—	973
LMArena Chinese	—	959
LMArena German	—	1014
LMArena Russian	—	941

Instruction Following Not comparable

Codestral: —, Llama 3.2 1B: 52.4 (#290)

Instruction Following benchmarks
Benchmark	Codestral	Llama 3.2 1B
LMArena Instruction Following	—	1031

Long Context Not comparable

Codestral: —, Llama 3.2 1B: 31.9 (#274)

Long Context benchmarks
Benchmark	Codestral	Llama 3.2 1B
LMArena Longer Query	—	1050

Writing & Preference Not comparable

Codestral: —, Llama 3.2 1B: 21.3 (#310)

Writing & Preference benchmarks
Benchmark	Codestral	Llama 3.2 1B
LMArena Text	—	1055
LMArena Creative Writing	—	1033
EQ-Bench Creative Writing	—	200
LMArena Multi-Turn	—	1030

Frequently asked questions

Is Codestral better than Llama 3.2 1B?

Which is cheaper, Codestral or Llama 3.2 1B?

Llama 3.2 1B is cheaper. It lists at $0.027 per million input tokens and $0.20 per million output tokens; Codestral lists at $0.30 and $0.90.

Is Codestral or Llama 3.2 1B better for coding?

Codestral scores higher on coding benchmarks: 27.3 versus 21.1 in the Noometry coding category.

Which has the bigger context window?

Codestral does, with 256K tokens against 60K.

How many benchmarks do Codestral and Llama 3.2 1B share?

2 benchmarks have published results for both models. Codestral has 7 scored results on Noometry and Llama 3.2 1B has 22.