Noometry

Model comparison

Amazon Nova Lite vs Devstral Small 2505

Devstral Small 2505 is the stronger model overall, scoring 33.9 to 31.4 on the Noometry Index.

Last verified . 0 shared benchmarks.

Amazon Nova Lite Amazon

31.4

Rank #262 Confirmed

Devstral Small 2505 Mistral AI

33.9

Rank #228 Reported

Summary

  • The widest gap is in coding, where Devstral Small 2505 leads 38.6 to 32.0.
  • Amazon Nova Lite is cheaper at $0.06 / $0.24 per million input/output tokens, against $0.10 / $0.30 for Devstral Small 2505.
  • Amazon Nova Lite accepts more context: 300K tokens versus 128K.
  • Devstral Small 2505 has downloadable open weights; the other is API-only.

Side by side

Amazon Nova LiteDevstral Small 2505
ProviderAmazonMistral AI
Noometry Index31.433.9
Released2024-12-032025-05-07
WeightsProprietaryOpen
Context window300K128K
Max output10K128K
Input $ / M tokens$0.06$0.10
Output $ / M tokens$0.24$0.30
Results tracked344

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

Category by category

Coding Devstral Small 2505 leads

Amazon Nova Lite: 32.0 (#266), Devstral Small 2505: 38.6 (#160)

BenchmarkAmazon Nova LiteDevstral Small 2505
SWE-bench Verified (bash only)—56.4%
SciCode—28.8%
LiveBench Coding27.5%—
LMArena Coding1239—
ALE-Bench236.25—

Reasoning Too close to call

Amazon Nova Lite: 18.5 (#257), Devstral Small 2505: 19.1 (#243)

BenchmarkAmazon Nova LiteDevstral Small 2505
Kagi LLM Benchmark—37.7%
CritPt—0%
LiveBench Reasoning36.7%—
LMArena Hard Prompts1220—
LiveBench Data Analysis37.2%—
LiveBench36.4%—

Math Not comparable

Amazon Nova Lite: 25.8 (#243), Devstral Small 2505: —

BenchmarkAmazon Nova LiteDevstral Small 2505
Omni-MATH23.3%—
LiveBench Math36.7%—
LMArena Math1227—

Knowledge Not comparable

Amazon Nova Lite: 26.0 (#254), Devstral Small 2505: —

BenchmarkAmazon Nova LiteDevstral Small 2505
Humanity's Last Exam3.6%—
MMLU-Pro60%—
Vectara Hallucination Rate6.1%—
GPQA (HELM)39.7%—
LMArena Expert1201—
MMLU77%—

Multimodal Not comparable

Amazon Nova Lite: 25.0 (#118), Devstral Small 2505: —

BenchmarkAmazon Nova LiteDevstral Small 2505
LMArena Vision990—

Multilingual Not comparable

Amazon Nova Lite: 38.2 (#227), Devstral Small 2505: —

BenchmarkAmazon Nova LiteDevstral Small 2505
LMArena Non-English1208—
LMArena Chinese1225—
LMArena French1237—
LMArena German1229—
LMArena Japanese1153—
LMArena Korean1153—
LMArena Russian1216—
LMArena Spanish1225—

Instruction Following Not comparable

Amazon Nova Lite: 59.6 (#250), Devstral Small 2505: —

BenchmarkAmazon Nova LiteDevstral Small 2505
LiveBench Instruction Following54.1%—
IFEval77.6%—
LMArena Instruction Following1205—

Long Context Not comparable

Amazon Nova Lite: 37.4 (#212), Devstral Small 2505: —

BenchmarkAmazon Nova LiteDevstral Small 2505
LMArena Longer Query1234—

Writing & Preference Not comparable

Amazon Nova Lite: 42.3 (#232), Devstral Small 2505: —

BenchmarkAmazon Nova LiteDevstral Small 2505
LMArena Text1229—
LMArena Creative Writing1197—
WildBench75%—
LMArena Multi-Turn1199—
LiveBench Language25.9%—

Frequently asked questions

Is Amazon Nova Lite better than Devstral Small 2505?

Devstral Small 2505 is the stronger model overall, scoring 33.9 to 31.4 on the Noometry Index.

Which is cheaper, Amazon Nova Lite or Devstral Small 2505?

Amazon Nova Lite is cheaper. It lists at $0.06 per million input tokens and $0.24 per million output tokens; Devstral Small 2505 lists at $0.10 and $0.30.

Is Amazon Nova Lite or Devstral Small 2505 better for coding?

Devstral Small 2505 scores higher on coding benchmarks: 38.6 versus 32.0 in the Noometry coding category.

Which has the bigger context window?

Amazon Nova Lite does, with 300K tokens against 128K.

How many benchmarks do Amazon Nova Lite and Devstral Small 2505 share?

0 benchmarks have published results for both models. Amazon Nova Lite has 34 scored results on Noometry and Devstral Small 2505 has 4.

Related comparisons

Go deeper