Noometry

Model comparison

Claude Haiku 4.5 vs Claude Opus 5

Claude Opus 5 is the stronger model overall, scoring 67.8 to 39.5 on the Noometry Index. Claude Haiku 4.5 costs 5.0× less per token, which makes it the better buy when Claude Opus 5's lead doesn't matter for your workload.

Last verified . 37 shared benchmarks.

Claude Haiku 4.5 Anthropic

39.5

Rank #165 Confirmed

Claude Opus 5 Anthropic

67.8

Rank #4 Confirmed

Summary

  • They share 37 benchmarks with published results for both. Claude Haiku 4.5 scores higher in 0 categories and Claude Opus 5 in 10 categories; 10 gaps are clear of the uncertainty.
  • The widest gap is in reasoning, where Claude Opus 5 leads 77.2 to 15.1.
  • The biggest single-benchmark swing is ARC-AGI-2: 4% for Claude Haiku 4.5 and 90.4% for Claude Opus 5.
  • Claude Haiku 4.5 is cheaper at $1 / $5 per million input/output tokens, against $5 / $25 for Claude Opus 5.
  • Claude Opus 5 accepts more context: 1M tokens versus 200K.

Side by side

Claude Haiku 4.5Claude Opus 5
ProviderAnthropicAnthropic
Noometry Index39.567.8
Released2025-10-152026-07-24
WeightsProprietaryProprietary
Context window200K1M
Max output64K128K
Input $ / M tokens$1$5
Output $ / M tokens$5$25
Results tracked5357

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

Category by category

Coding Claude Opus 5 leads

Claude Haiku 4.5: 44.0 (#78), Claude Opus 5: 67.5 (#5)

BenchmarkClaude Haiku 4.5Claude Opus 5
LMArena WebDev13301691
SciCode43.3%56.4%
WeirdML45.4%91.8%
LMArena Coding14531534
ALE-Bench653.482,165
DeepSWE—73.6%
FrontierCode—53.4%
SWE-bench Verified (bash only)66.6%—
CursorBench—46.6%
SWE-bench Multilingual64.7%—
FrontierSWE—52%

Agentic & Tool Use Claude Opus 5 leads

Claude Haiku 4.5: 33.6 (#52), Claude Opus 5: 55.6 (#1)

BenchmarkClaude Haiku 4.5Claude Opus 5
BALROG31.2%63.4%
Vending-Bench 2458.8911,182
Terminal-Bench35.5%—
APEX-Agents—65.8%
Berkeley Function Calling Leaderboard68.7%—
OSWorld 2.0—31.4%
τ²-bench Banking—48.7%
DeepResearch Bench45.5%—
PostTrainBench—35%
ExploitBench13.7%—
GBAEval—79.6%
GDP.pdf—24%

Reasoning Claude Opus 5 leads

Claude Haiku 4.5: 15.1 (#320), Claude Opus 5: 77.2 (#4)

BenchmarkClaude Haiku 4.5Claude Opus 5
ARC-AGI-24%90.4%
NYT Connections (extended)14.3%94.3%
ARC-AGI-147.7%97.5%
CritPt0%29.1%
Chess Puzzles8%42%
LMArena Hard Prompts14201526
DTBench73.6%97.9%
LMCA30.9%64.5%
Epoch Capabilities Index142.41162.78
SimpleBench—80.6%
EBR-Bench—45.7%
Mystery Game Puzzles—59%
Bench to the Future 3—0.12
ForecastBench61.4—

Math Claude Opus 5 leads

Claude Haiku 4.5: 44.9 (#78), Claude Opus 5: 86.2 (#8)

BenchmarkClaude Haiku 4.5Claude Opus 5
OTIS Mock AIME 2024-202566.7%98.9%
LMArena Math13961531
FrontierMath (Tiers 1-3)—85.6%
FrontierMath Tier 4—73.2%
ProofBench—99%
Omni-MATH56.1%—
MATH Level 596.4%—
FrontierMath (Feb 2025 set)5.9%—
FrontierMath Tier 4 (v1)2.1%—

Knowledge Claude Opus 5 leads

Claude Haiku 4.5: 37.7 (#153), Claude Opus 5: 66.8 (#9)

BenchmarkClaude Haiku 4.5Claude Opus 5
GPQA Diamond71.2%93.9%
SimpleQA Verified13.2%59.9%
LMArena Expert14421557
MMLU-Pro77.7%—
Vectara Hallucination Rate9.8%—
GPQA (HELM)60.5%—

Multimodal Claude Opus 5 leads

Claude Haiku 4.5: 26.8 (#118), Claude Opus 5: 50.8 (#8)

BenchmarkClaude Haiku 4.5Claude Opus 5
Blueprint-Bench 20%30.4%
LMArena Document14201516
LMArena Vision—1319
Furniture Assembly—60.8%

Multilingual Claude Opus 5 leads

Claude Haiku 4.5: 49.9 (#129), Claude Opus 5: 58.8 (#4)

BenchmarkClaude Haiku 4.5Claude Opus 5
LMArena Non-English13771501
LMArena Chinese14171574
LMArena French14081519
LMArena German13751524
LMArena Japanese13391516
LMArena Korean13471521
LMArena Russian13811507
LMArena Spanish14201519

Instruction Following Claude Opus 5 leads

Claude Haiku 4.5: 71.4 (#149), Claude Opus 5: 79.2 (#7)

BenchmarkClaude Haiku 4.5Claude Opus 5
LMArena Instruction Following14141517
IFEval80.1%—

Long Context Claude Opus 5 leads

Claude Haiku 4.5: 43.6 (#92), Claude Opus 5: 46.5 (#21)

BenchmarkClaude Haiku 4.5Claude Opus 5
LMArena Longer Query14271515

Writing & Preference Claude Opus 5 leads

Claude Haiku 4.5: 57.9 (#123), Claude Opus 5: 79.2 (#1)

BenchmarkClaude Haiku 4.5Claude Opus 5
LMArena Text13961507
LMArena Creative Writing13721491
EQ-Bench 410641385
LMArena Multi-Turn14091499
EQ-Bench Creative Writing—2133
WildBench83.9%—

Frequently asked questions

Is Claude Haiku 4.5 better than Claude Opus 5?

Claude Opus 5 is the stronger model overall, scoring 67.8 to 39.5 on the Noometry Index. Claude Haiku 4.5 costs 5.0× less per token, which makes it the better buy when Claude Opus 5's lead doesn't matter for your workload.

Which is cheaper, Claude Haiku 4.5 or Claude Opus 5?

Claude Haiku 4.5 is cheaper. It lists at $1 per million input tokens and $5 per million output tokens; Claude Opus 5 lists at $5 and $25.

Is Claude Haiku 4.5 or Claude Opus 5 better for coding?

Claude Opus 5 scores higher on coding benchmarks: 67.5 versus 44.0 in the Noometry coding category.

Which has the bigger context window?

Claude Opus 5 does, with 1M tokens against 200K.

How many benchmarks do Claude Haiku 4.5 and Claude Opus 5 share?

37 benchmarks have published results for both models. Claude Haiku 4.5 has 53 scored results on Noometry and Claude Opus 5 has 57.

Related comparisons

Go deeper