Model comparison
Amazon Nova Lite vs Claude Haiku 4.5
Claude Haiku 4.5 is the stronger model overall, scoring 39.5 to 31.9 on the Noometry Index. Amazon Nova Lite costs 19× less per token, which makes it the better buy when Claude Haiku 4.5's lead doesn't matter for your workload.
Last verified . 24 shared benchmarks.
Summary
- They share 24 benchmarks with published results for both. Amazon Nova Lite scores higher in 1 category and Claude Haiku 4.5 in 8 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where Claude Haiku 4.5 leads 44.9 to 27.8.
- The biggest single-benchmark swing is Omni-MATH: 23.3% for Amazon Nova Lite and 56.1% for Claude Haiku 4.5.
- Amazon Nova Lite is cheaper at $0.06 / $0.24 per million input/output tokens, against $1 / $5 for Claude Haiku 4.5.
- Amazon Nova Lite accepts more context: 300K tokens versus 200K.
Side by side
| Amazon Nova Lite | Claude Haiku 4.5 | |
|---|---|---|
| Provider | Amazon | Anthropic |
| Noometry Index | 31.9 | 39.5 |
| Released | 2024-12-03 | 2025-10-15 |
| Weights | Proprietary | Proprietary |
| Context window | 300K | 200K |
| Max output | 10K | 64K |
| Input $ / M tokens | $0.06 | $1 |
| Output $ / M tokens | $0.24 | $5 |
| Results tracked | 34 | 53 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Claude Haiku 4.5 leads
Amazon Nova Lite: 32.5 (#270), Claude Haiku 4.5: 44.0 (#78)
| Benchmark | Amazon Nova Lite | Claude Haiku 4.5 |
|---|---|---|
| LMArena Coding | 1239 | 1453 |
| ALE-Bench | 236.25 | 653.48 |
| SWE-bench Verified (bash only) | — | 66.6% |
| LMArena WebDev | — | 1330 |
| SWE-bench Multilingual | — | 64.7% |
| SciCode | — | 43.3% |
| WeirdML | — | 45.4% |
| LiveBench Coding | 27.5% | — |
Agentic & Tool Use Not comparable
Amazon Nova Lite: —, Claude Haiku 4.5: 33.6 (#52)
| Benchmark | Amazon Nova Lite | Claude Haiku 4.5 |
|---|---|---|
| Terminal-Bench | — | 35.5% |
| Berkeley Function Calling Leaderboard | — | 68.7% |
| DeepResearch Bench | — | 45.5% |
| BALROG | — | 31.2% |
| ExploitBench | — | 13.7% |
| Vending-Bench 2 | — | 458.89 |
Reasoning Amazon Nova Lite leads
Amazon Nova Lite: 19.3 (#260), Claude Haiku 4.5: 15.1 (#320)
| Benchmark | Amazon Nova Lite | Claude Haiku 4.5 |
|---|---|---|
| LMArena Hard Prompts | 1220 | 1420 |
| ARC-AGI-2 | — | 4% |
| NYT Connections (extended) | — | 14.3% |
| ARC-AGI-1 | — | 47.7% |
| CritPt | — | 0% |
| Chess Puzzles | — | 8% |
| LiveBench Reasoning | 36.7% | — |
| DTBench | — | 73.6% |
| LiveBench Data Analysis | 37.2% | — |
| LMCA | — | 30.9% |
| Epoch Capabilities Index | — | 142.41 |
| ForecastBench | — | 61.4 |
| LiveBench | 36.4% | — |
Math Claude Haiku 4.5 leads
Amazon Nova Lite: 27.8 (#247), Claude Haiku 4.5: 44.9 (#78)
| Benchmark | Amazon Nova Lite | Claude Haiku 4.5 |
|---|---|---|
| Omni-MATH | 23.3% | 56.1% |
| LMArena Math | 1227 | 1396 |
| OTIS Mock AIME 2024-2025 | — | 66.7% |
| LiveBench Math | 36.7% | — |
| MATH Level 5 | — | 96.4% |
| FrontierMath (Feb 2025 set) | — | 5.9% |
| FrontierMath Tier 4 (v1) | — | 2.1% |
Knowledge Claude Haiku 4.5 leads
Amazon Nova Lite: 26.3 (#259), Claude Haiku 4.5: 37.7 (#153)
| Benchmark | Amazon Nova Lite | Claude Haiku 4.5 |
|---|---|---|
| MMLU-Pro | 60% | 77.7% |
| Vectara Hallucination Rate | 6.1% | 9.8% |
| GPQA (HELM) | 39.7% | 60.5% |
| LMArena Expert | 1201 | 1442 |
| GPQA Diamond | — | 71.2% |
| Humanity's Last Exam | 3.6% | — |
| SimpleQA Verified | — | 13.2% |
| MMLU | 77% | — |
Multimodal Claude Haiku 4.5 leads
Amazon Nova Lite: 25.5 (#123), Claude Haiku 4.5: 26.8 (#118)
| Benchmark | Amazon Nova Lite | Claude Haiku 4.5 |
|---|---|---|
| LMArena Vision | 990 | — |
| Blueprint-Bench 2 | — | 0% |
| LMArena Document | — | 1420 |
Multilingual Claude Haiku 4.5 leads
Amazon Nova Lite: 38.0 (#232), Claude Haiku 4.5: 49.9 (#129)
| Benchmark | Amazon Nova Lite | Claude Haiku 4.5 |
|---|---|---|
| LMArena Non-English | 1208 | 1377 |
| LMArena Chinese | 1225 | 1417 |
| LMArena French | 1237 | 1408 |
| LMArena German | 1229 | 1375 |
| LMArena Japanese | 1153 | 1339 |
| LMArena Korean | 1153 | 1347 |
| LMArena Russian | 1216 | 1381 |
| LMArena Spanish | 1226 | 1420 |
Instruction Following Claude Haiku 4.5 leads
Amazon Nova Lite: 59.3 (#255), Claude Haiku 4.5: 71.4 (#149)
| Benchmark | Amazon Nova Lite | Claude Haiku 4.5 |
|---|---|---|
| IFEval | 77.6% | 80.1% |
| LMArena Instruction Following | 1205 | 1414 |
| LiveBench Instruction Following | 54.1% | — |
Long Context Claude Haiku 4.5 leads
Amazon Nova Lite: 37.4 (#217), Claude Haiku 4.5: 43.6 (#92)
| Benchmark | Amazon Nova Lite | Claude Haiku 4.5 |
|---|---|---|
| LMArena Longer Query | 1234 | 1427 |
Writing & Preference Claude Haiku 4.5 leads
Amazon Nova Lite: 42.1 (#237), Claude Haiku 4.5: 57.9 (#123)
| Benchmark | Amazon Nova Lite | Claude Haiku 4.5 |
|---|---|---|
| LMArena Text | 1229 | 1396 |
| LMArena Creative Writing | 1197 | 1372 |
| WildBench | 75% | 83.9% |
| LMArena Multi-Turn | 1199 | 1409 |
| EQ-Bench 4 | — | 1064 |
| LiveBench Language | 25.9% | — |
Frequently asked questions
Is Amazon Nova Lite better than Claude Haiku 4.5?
Claude Haiku 4.5 is the stronger model overall, scoring 39.5 to 31.9 on the Noometry Index. Amazon Nova Lite costs 19× 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, Amazon Nova Lite or Claude Haiku 4.5?
Amazon Nova Lite is cheaper. It lists at $0.06 per million input tokens and $0.24 per million output tokens; Claude Haiku 4.5 lists at $1 and $5.
Is Amazon Nova Lite or Claude Haiku 4.5 better for coding?
Claude Haiku 4.5 scores higher on coding benchmarks: 44.0 versus 32.5 in the Noometry coding category.
Which has the bigger context window?
Amazon Nova Lite does, with 300K tokens against 200K.
How many benchmarks do Amazon Nova Lite and Claude Haiku 4.5 share?
24 benchmarks have published results for both models. Amazon Nova Lite has 34 scored results on Noometry and Claude Haiku 4.5 has 53.