Noometry

Model comparison

Claude Haiku 4.5 vs Codestral

Claude Haiku 4.5 is the stronger model overall, scoring 39.3 to 30.1 on the Noometry Index. Codestral costs 4.4× less per token, which makes it the better buy when Claude Haiku 4.5's lead doesn't matter for your workload.

Last verified . 1 shared benchmarks.

Claude Haiku 4.5 Anthropic

39.3

Rank #155 Confirmed

Codestral Mistral AI

30.1

Rank #287 Reported

Summary

  • They share 1 benchmark with published results for both. Claude Haiku 4.5 scores higher in 1 category and Codestral in 1 category; 2 gaps are clear of the uncertainty.
  • The widest gap is in coding, where Claude Haiku 4.5 leads 43.9 to 26.9.
  • Codestral is cheaper at $0.30 / $0.90 per million input/output tokens, against $1 / $5 for Claude Haiku 4.5.
  • Codestral accepts more context: 256K tokens versus 200K.
  • Codestral has downloadable open weights; the other is API-only.

Side by side

Claude Haiku 4.5Codestral
ProviderAnthropicMistral AI
Noometry Index39.330.1
Released2025-10-152024-05-29
WeightsProprietaryOpen
Context window200K256K
Max output64K4K
Input $ / M tokens$1$0.30
Output $ / M tokens$5$0.90
Results tracked537

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

Category by category

Coding Claude Haiku 4.5 leads

Claude Haiku 4.5: 43.9 (#71), Codestral: 26.9 (#316)

BenchmarkClaude Haiku 4.5Codestral
ALE-Bench653.48137.78
SWE-bench Verified (bash only)66.6%—
Aider Polyglot—11.1%
LMArena WebDev1330—
SWE-bench Multilingual64.7%—
SciCode43.3%—
WeirdML45.4%—
BigCodeBench Instruct—41.8%
LMArena Coding1454—
BigCodeBench Complete—52.5%
HumanEval+—73.8%
MBPP+—61.9%

Agentic & Tool Use Not comparable

Claude Haiku 4.5: 33.5 (#50), Codestral: —

BenchmarkClaude Haiku 4.5Codestral
Terminal-Bench35.5%—
Berkeley Function Calling Leaderboard68.7%—
DeepResearch Bench45.5%—
BALROG31.2%—
ExploitBench13.7%—
Vending-Bench 2458.89—

Reasoning Codestral leads

Claude Haiku 4.5: 15.1 (#310), Codestral: 19.1 (#244)

BenchmarkClaude Haiku 4.5Codestral
ARC-AGI-24%—
Kagi LLM Benchmark—32.5%
NYT Connections (extended)14.3%—
ARC-AGI-147.7%—
CritPt0%—
Chess Puzzles8%—
LMArena Hard Prompts1420—
DTBench73.6%—
LMCA30.9%—
Epoch Capabilities Index142.42—
ForecastBench61.4—

Math Not comparable

Claude Haiku 4.5: 43.0 (#69), Codestral: —

BenchmarkClaude Haiku 4.5Codestral
OTIS Mock AIME 2024-202566.7%—
Omni-MATH56.1%—
LMArena Math1395—
MATH Level 596.4%—
FrontierMath (Feb 2025 set)5.9%—
FrontierMath Tier 4 (v1)2.1%—

Knowledge Not comparable

Claude Haiku 4.5: 37.6 (#146), Codestral: —

BenchmarkClaude Haiku 4.5Codestral
GPQA Diamond71.2%—
SimpleQA Verified13.2%—
MMLU-Pro77.7%—
Vectara Hallucination Rate9.8%—
GPQA (HELM)60.5%—
LMArena Expert1443—

Multimodal Not comparable

Claude Haiku 4.5: 27.0 (#111), Codestral: —

BenchmarkClaude Haiku 4.5Codestral
Blueprint-Bench 20%—
LMArena Document1420—

Multilingual Not comparable

Claude Haiku 4.5: 50.1 (#124), Codestral: —

BenchmarkClaude Haiku 4.5Codestral
LMArena Non-English1376—
LMArena Chinese1415—
LMArena French1408—
LMArena German1377—
LMArena Japanese1339—
LMArena Korean1346—
LMArena Russian1380—
LMArena Spanish1420—

Instruction Following Not comparable

Claude Haiku 4.5: 71.6 (#144), Codestral: —

BenchmarkClaude Haiku 4.5Codestral
IFEval80.1%—
LMArena Instruction Following1414—

Long Context Not comparable

Claude Haiku 4.5: 43.6 (#86), Codestral: —

BenchmarkClaude Haiku 4.5Codestral
LMArena Longer Query1427—

Writing & Preference Not comparable

Claude Haiku 4.5: 58.1 (#119), Codestral: —

BenchmarkClaude Haiku 4.5Codestral
LMArena Text1396—
LMArena Creative Writing1371—
WildBench83.9%—
EQ-Bench 41064—
LMArena Multi-Turn1409—

Frequently asked questions

Is Claude Haiku 4.5 better than Codestral?

Claude Haiku 4.5 is the stronger model overall, scoring 39.3 to 30.1 on the Noometry Index. Codestral costs 4.4× less per token, which makes it the better buy when Claude Haiku 4.5's lead doesn't matter for your workload.

Which is cheaper, Claude Haiku 4.5 or Codestral?

Codestral is cheaper. It lists at $0.30 per million input tokens and $0.90 per million output tokens; Claude Haiku 4.5 lists at $1 and $5.

Is Claude Haiku 4.5 or Codestral better for coding?

Claude Haiku 4.5 scores higher on coding benchmarks: 43.9 versus 26.9 in the Noometry coding category.

Which has the bigger context window?

Codestral does, with 256K tokens against 200K.

How many benchmarks do Claude Haiku 4.5 and Codestral share?

1 benchmark has published results for both models. Claude Haiku 4.5 has 53 scored results on Noometry and Codestral has 7.

Related comparisons

Go deeper