Rankings / Amazon
Nova Lite
released 2024-12-03
BenchAtlas Index
as of 2026-07-14
base configuration
rank #423
11 families · 6 categories · high
Benchmark evidence
15 results
agentic coding
| τ²-Bench | 17.5 % | independent | Artificial Analysis |
coding
| LiveCodeBench v6 implementation=artificial-analysis | 16.7 % | independent | Artificial Analysis | |
| SciCode | 13.9 % | independent | Artificial Analysis |
external indices
| Intelligence Index v4.1 | 6.9 points | independent | Artificial Analysis | |
| AA Math Index | 7.0 points | independent | Artificial Analysis |
knowledge science
| GPQA Diamond implementation=artificial-analysis | 43.3 % | independent | Artificial Analysis | |
| Humanity's Last Exam implementation=artificial-analysis | 4.6 % | independent | Artificial Analysis | |
| Humanity's Last Exam contamination=Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions. · implementation=scale | 3.6 % 2.9–4.4 | independent | Scale Labs | |
| MMLU-Pro implementation=artificial-analysis | 59.0 % | independent | Artificial Analysis |
long context instruction
| AA-LCR | 17.7 % | independent | Artificial Analysis | |
| IFBench | 34.1 % | independent | Artificial Analysis |
multimodal
| VISTA | 25.5 % 24.7–26.3 | independent | Scale Labs |
reasoning math
| AIME implementation=artificial-analysis | 10.7 % | independent | Artificial Analysis | |
| AIME year=2025 · implementation=artificial-analysis | 7.0 % | independent | Artificial Analysis | |
| MATH-500 implementation=artificial-analysis | 76.5 % | independent | Artificial Analysis |
Agent + model results
systems, not bare-model scores
| agent + model Artificial Analysis harness + Nova Lite | Terminal-Bench Hard | 0.8 % | independent | Artificial Analysis |
These scores measure the whole agent system (scaffold, tools, budgets) — they are never merged into the bare model’s numbers.