Noometry

Model comparison

Claude Haiku 4.5 vs Step 3

Claude Haiku 4.5 and Step 3 score almost the same on the Noometry Index (39.5 vs 40.5), so choose on price, context window or the category you care about most.

Last verified . 15 shared benchmarks.

Claude Haiku 4.5 Anthropic

39.5

Rank #165 Confirmed

Step 3 StepFun

40.5

Rank #149 Confirmed

Summary

  • They share 15 benchmarks with published results for both. Claude Haiku 4.5 scores higher in 7 categories and Step 3 in 2 categories; 8 gaps are clear of the uncertainty.
  • The widest gap is in reasoning, where Step 3 leads 28.4 to 15.1.
  • Step 3 has downloadable open weights; the other is API-only.

Side by side

Claude Haiku 4.5Step 3
ProviderAnthropicStepFun
Noometry Index39.540.5
Released2025-10-15—
WeightsProprietaryOpen
Context window200K—
Max output64K—
Input $ / M tokens$1—
Output $ / M tokens$5—
Results tracked5317

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: 44.0 (#78), Step 3: 40.1 (#147)

BenchmarkClaude Haiku 4.5Step 3
LMArena Coding14531367
SWE-bench Verified (bash only)66.6%—
LMArena WebDev1330—
SWE-bench Multilingual64.7%—
SciCode43.3%—
WeirdML45.4%—
ALE-Bench653.48—

Agentic & Tool Use Not comparable

Claude Haiku 4.5: 33.6 (#52), Step 3: —

BenchmarkClaude Haiku 4.5Step 3
Terminal-Bench35.5%—
Berkeley Function Calling Leaderboard68.7%—
DeepResearch Bench45.5%—
BALROG31.2%—
ExploitBench13.7%—
Vending-Bench 2458.89—

Reasoning Step 3 leads

Claude Haiku 4.5: 15.1 (#320), Step 3: 28.4 (#105)

BenchmarkClaude Haiku 4.5Step 3
LMArena Hard Prompts14201355
ARC-AGI-24%—
Kagi LLM Benchmark—62.3%
NYT Connections (extended)14.3%—
ARC-AGI-147.7%—
CritPt0%—
Chess Puzzles8%—
DTBench73.6%—
LMCA30.9%—
Epoch Capabilities Index142.41—
ForecastBench61.4—

Math Claude Haiku 4.5 leads

Claude Haiku 4.5: 44.9 (#78), Step 3: 37.6 (#148)

BenchmarkClaude Haiku 4.5Step 3
LMArena Math13961366
OTIS Mock AIME 2024-202566.7%—
Omni-MATH56.1%—
MATH Level 596.4%—
FrontierMath (Feb 2025 set)5.9%—
FrontierMath Tier 4 (v1)2.1%—

Knowledge Too close to call

Claude Haiku 4.5: 37.7 (#153), Step 3: 36.8 (#164)

BenchmarkClaude Haiku 4.5Step 3
LMArena Expert14421333
GPQA Diamond71.2%—
SimpleQA Verified13.2%—
MMLU-Pro77.7%—
Vectara Hallucination Rate9.8%—
GPQA (HELM)60.5%—

Multimodal Step 3 leads

Claude Haiku 4.5: 26.8 (#118), Step 3: 35.5 (#86)

BenchmarkClaude Haiku 4.5Step 3
LMArena Vision—1177
Blueprint-Bench 20%—
LMArena Document1420—

Multilingual Claude Haiku 4.5 leads

Claude Haiku 4.5: 49.9 (#129), Step 3: 46.3 (#159)

BenchmarkClaude Haiku 4.5Step 3
LMArena Non-English13771327
LMArena Chinese14171397
LMArena German13751371
LMArena Korean13471269
LMArena Russian13811331
LMArena Spanish14201371
LMArena French1408—
LMArena Japanese1339—

Instruction Following Claude Haiku 4.5 leads

Claude Haiku 4.5: 71.4 (#149), Step 3: 70.4 (#164)

BenchmarkClaude Haiku 4.5Step 3
LMArena Instruction Following14141332
IFEval80.1%—

Long Context Claude Haiku 4.5 leads

Claude Haiku 4.5: 43.6 (#92), Step 3: 40.3 (#157)

BenchmarkClaude Haiku 4.5Step 3
LMArena Longer Query14271326

Writing & Preference Claude Haiku 4.5 leads

Claude Haiku 4.5: 57.9 (#123), Step 3: 54.3 (#151)

BenchmarkClaude Haiku 4.5Step 3
LMArena Text13961350
LMArena Creative Writing13721321
LMArena Multi-Turn14091341
WildBench83.9%—
EQ-Bench 41064—

Frequently asked questions

Is Claude Haiku 4.5 better than Step 3?

Claude Haiku 4.5 and Step 3 score almost the same on the Noometry Index (39.5 vs 40.5), so choose on price, context window or the category you care about most.

Is Claude Haiku 4.5 or Step 3 better for coding?

Claude Haiku 4.5 scores higher on coding benchmarks: 44.0 versus 40.1 in the Noometry coding category.

How many benchmarks do Claude Haiku 4.5 and Step 3 share?

15 benchmarks have published results for both models. Claude Haiku 4.5 has 53 scored results on Noometry and Step 3 has 17.

Related comparisons

Go deeper