Noometry

Model comparison

Claude 3 Haiku vs Qwen2.5-Coder-32B

Qwen2.5-Coder-32B is the stronger model overall, scoring 33.4 to 25.9 on the Noometry Index.

Last verified . 19 shared benchmarks.

Claude 3 Haiku Anthropic

25.9

Rank #340 Confirmed

Qwen2.5-Coder-32B Alibaba (Qwen)

33.4

Rank #245 Confirmed

Summary

  • They share 19 benchmarks with published results for both. Claude 3 Haiku scores higher in 1 category and Qwen2.5-Coder-32B in 7 categories; 7 gaps are clear of the uncertainty.
  • The widest gap is in math, where Qwen2.5-Coder-32B leads 33.3 to 9.8.
  • The biggest single-benchmark swing is BigCodeBench Instruct: 39.4% for Claude 3 Haiku and 49% for Qwen2.5-Coder-32B.
  • Qwen2.5-Coder-32B has downloadable open weights; the other is API-only.

Side by side

Claude 3 HaikuQwen2.5-Coder-32B
ProviderAnthropicAlibaba (Qwen)
Noometry Index25.933.4
Released2024-03-072024-09-18
WeightsProprietaryOpen
Context window—33K
Max output—29K
Input $ / M tokens—$0.66
Output $ / M tokens—$1
Results tracked3731

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

Category by category

Coding Claude 3 Haiku leads

Claude 3 Haiku: 26.4 (#325), Qwen2.5-Coder-32B: 22.6 (#333)

BenchmarkClaude 3 HaikuQwen2.5-Coder-32B
BigCodeBench Instruct39.4%49%
LMArena Coding11991276
BigCodeBench Complete50.1%58%
HumanEval+68.9%87.2%
MBPP+68.8%77%
SWE-bench Verified (bash only)—9%
Aider Polyglot—16.4%
WeirdML9.8%—
LiveBench Coding—56.9%
CadEval12%—

Reasoning Qwen2.5-Coder-32B leads

Claude 3 Haiku: 16.3 (#307), Qwen2.5-Coder-32B: 21.2 (#225)

BenchmarkClaude 3 HaikuQwen2.5-Coder-32B
LMArena Hard Prompts11741251
Epoch Capabilities Index118.35119.49
WinoGrande74.2%80.8%
Kagi LLM Benchmark34.2%—
LiveBench Reasoning—42.1%
DTBench50.1%—
LiveBench Data Analysis—49.9%
LMCA8.8%—
ForecastBench53.2—
HellaSwag—83%
LiveBench—46.2%

Math Qwen2.5-Coder-32B leads

Claude 3 Haiku: 9.8 (#319), Qwen2.5-Coder-32B: 33.3 (#204)

BenchmarkClaude 3 HaikuQwen2.5-Coder-32B
LMArena Math11881251
OTIS Mock AIME 2024-20251.8%—
LiveBench Math—46.6%
MATH Level 514.9%—
GSM8K—93%

Knowledge Qwen2.5-Coder-32B leads

Claude 3 Haiku: 17.3 (#285), Qwen2.5-Coder-32B: 33.4 (#203)

BenchmarkClaude 3 HaikuQwen2.5-Coder-32B
LMArena Expert11481221
MMLU73.8%79.1%
GPQA Diamond36.3%—
Confabulations34.2%—
ARC (AI2) Challenge—70.5%

Multimodal Not comparable

Claude 3 Haiku: 23.6 (#128), Qwen2.5-Coder-32B: —

BenchmarkClaude 3 HaikuQwen2.5-Coder-32B
LMArena Vision950—
ScienceQA72%—

Multilingual Qwen2.5-Coder-32B leads

Claude 3 Haiku: 36.0 (#243), Qwen2.5-Coder-32B: 37.8 (#235)

BenchmarkClaude 3 HaikuQwen2.5-Coder-32B
LMArena Non-English11781205
LMArena Chinese11551222
LMArena Russian12041228
LMArena French1195—
LMArena German1174—
LMArena Japanese1102—
LMArena Korean1109—
LMArena Spanish1166—

Instruction Following Too close to call

Claude 3 Haiku: 61.3 (#247), Qwen2.5-Coder-32B: 61.4 (#245)

BenchmarkClaude 3 HaikuQwen2.5-Coder-32B
LMArena Instruction Following11731223
LiveBench Instruction Following—58.7%

Long Context Qwen2.5-Coder-32B leads

Claude 3 Haiku: 36.1 (#237), Qwen2.5-Coder-32B: 38.0 (#208)

BenchmarkClaude 3 HaikuQwen2.5-Coder-32B
LMArena Longer Query11901251

Writing & Preference Qwen2.5-Coder-32B leads

Claude 3 Haiku: 29.7 (#291), Qwen2.5-Coder-32B: 41.6 (#240)

BenchmarkClaude 3 HaikuQwen2.5-Coder-32B
LMArena Text11951230
LMArena Creative Writing11571174
LMArena Multi-Turn11901222
EQ-Bench Creative Writing717—
LiveBench Language—23.3%

Frequently asked questions

Is Claude 3 Haiku better than Qwen2.5-Coder-32B?

Qwen2.5-Coder-32B is the stronger model overall, scoring 33.4 to 25.9 on the Noometry Index.

Is Claude 3 Haiku or Qwen2.5-Coder-32B better for coding?

Claude 3 Haiku scores higher on coding benchmarks: 26.4 versus 22.6 in the Noometry coding category.

How many benchmarks do Claude 3 Haiku and Qwen2.5-Coder-32B share?

19 benchmarks have published results for both models. Claude 3 Haiku has 37 scored results on Noometry and Qwen2.5-Coder-32B has 31.

Related comparisons

Go deeper