Model comparison

Codestral vs Nova 2 Lite

Nova 2 Lite is the stronger model overall, scoring 39.7 to 30.6 on the Noometry Index. Codestral costs 1.9× less per token, which makes it the better buy when Nova 2 Lite's lead doesn't matter for your workload.

Last verified . 0 shared benchmarks.

Codestral Mistral AI

30.6

Rank #290 Reported

Nova 2 Lite Amazon

39.7

Rank #161 Confirmed

Summary

  • The widest gap is in coding, where Nova 2 Lite leads 40.7 to 27.3.
  • Codestral is cheaper at $0.30 / $0.90 per million input/output tokens, against $0.30 / $2.50 for Nova 2 Lite.
  • Nova 2 Lite accepts more context: 1M tokens versus 256K.

Side by side

Codestral and Nova 2 Lite specifications
CodestralNova 2 Lite
ProviderMistral AIAmazon
Noometry Index30.639.7
Released2024-05-292025-12-01
WeightsProprietaryProprietary
Context window256K1M
Max output8K64K
Input $ / M tokens$0.30$0.30
Output $ / M tokens$0.90$2.50
Results tracked719

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

Category by category

Coding Nova 2 Lite leads

Codestral: 27.3 (#321), Nova 2 Lite: 40.7 (#134)

Coding benchmarks
BenchmarkCodestralNova 2 Lite
Aider Polyglot11.1%—
BigCodeBench Instruct41.8%—
LMArena Coding—1385
BigCodeBench Complete52.5%—
ALE-Bench137.78—
HumanEval+73.8%—
MBPP+61.9%—

Agentic & Tool Use Not comparable

Codestral: —, Nova 2 Lite: 24.1 (#121)

Agentic & Tool Use benchmarks
BenchmarkCodestralNova 2 Lite
Berkeley Function Calling Leaderboard—27.1%

Reasoning Nova 2 Lite leads

Codestral: 19.8 (#251), Nova 2 Lite: 27.5 (#118)

Reasoning benchmarks
BenchmarkCodestralNova 2 Lite
Kagi LLM Benchmark32.5%—
LMArena Hard Prompts—1364

Math Not comparable

Codestral: —, Nova 2 Lite: 37.5 (#156)

Math benchmarks
BenchmarkCodestralNova 2 Lite
LMArena Math—1359

Knowledge Not comparable

Codestral: —, Nova 2 Lite: 43.0 (#94)

Knowledge benchmarks
BenchmarkCodestralNova 2 Lite
Vectara Hallucination Rate—5.1%
LMArena Expert—1358

Multilingual Not comparable

Codestral: —, Nova 2 Lite: 47.1 (#153)

Multilingual benchmarks
BenchmarkCodestralNova 2 Lite
LMArena Non-English—1337
LMArena Chinese—1364
LMArena French—1381
LMArena German—1343
LMArena Japanese—1271
LMArena Korean—1284
LMArena Russian—1343
LMArena Spanish—1373

Instruction Following Not comparable

Codestral: —, Nova 2 Lite: 70.5 (#161)

Instruction Following benchmarks
BenchmarkCodestralNova 2 Lite
LMArena Instruction Following—1335

Long Context Not comparable

Codestral: —, Nova 2 Lite: 40.6 (#150)

Long Context benchmarks
BenchmarkCodestralNova 2 Lite
LMArena Longer Query—1335

Writing & Preference Not comparable

Codestral: —, Nova 2 Lite: 53.9 (#154)

Writing & Preference benchmarks
BenchmarkCodestralNova 2 Lite
LMArena Text—1362
LMArena Creative Writing—1291
LMArena Multi-Turn—1338

Frequently asked questions

Is Codestral better than Nova 2 Lite?

Nova 2 Lite is the stronger model overall, scoring 39.7 to 30.6 on the Noometry Index. Codestral costs 1.9× less per token, which makes it the better buy when Nova 2 Lite's lead doesn't matter for your workload.

Which is cheaper, Codestral or Nova 2 Lite?

Codestral is cheaper. It lists at $0.30 per million input tokens and $0.90 per million output tokens; Nova 2 Lite lists at $0.30 and $2.50.

Is Codestral or Nova 2 Lite better for coding?

Nova 2 Lite scores higher on coding benchmarks: 40.7 versus 27.3 in the Noometry coding category.

Which has the bigger context window?

Nova 2 Lite does, with 1M tokens against 256K.

How many benchmarks do Codestral and Nova 2 Lite share?

0 benchmarks have published results for both models. Codestral has 7 scored results on Noometry and Nova 2 Lite has 19.

Related comparisons

Go deeper