Released: 2025-05-18

Amazon Nova 2 Lite

Second-generation Nova Lite architecture bringing expanded 500K context and upgraded visual artifact generation.

Intelligence Index
82.0/ 100
Calibrated multi-domain composite
Throughput Speed
175tok/sec
Output streaming throughput
Time To First Token
190ms
Initial chunk server latency
LiveBench Contam-Free
72.8%
September 2026 suite score

Sourced Benchmark Evaluations (61)

Verified performance across authoritative benchmarks with provenance tracking.
BenchmarkStatusScoreTrustDateSource TypeProvenance
ARC-AGI-3Active52.7%
93
2025-06-22independentSource ↗
ARC-AGI-2Nearing Saturation63.2%
75
2025-06-15independentSource ↗
FrontierMathActive22.3%
98
2025-06-12independentSource ↗
PutnamBenchActive41.3%
90
2025-06-11independentSource ↗
SWE-LancerDeprecated50.3%
20
2025-06-09independentSource ↗
Omni-MATHDeprecated58.8%
10
2025-06-09independentSource ↗
SoohakActive60.4%
98
2025-06-07independentSource ↗
FrontierCodeActive42.9%
98
2025-06-07independentSource ↗
Humanity's Last ExamActive35.6%
93
2025-06-05independentSource ↗
StructFlowBenchNearing Saturation64.1%
75
2025-06-05independentSource ↗
AIMENearing Saturation75.5%
75
2025-06-05independentSource ↗
OlympiadBenchActive65.2%
80
2025-06-04independentSource ↗
MathArenaActive63.6%
99
2025-06-03independentSource ↗
FireBenchActive63.4%
95
2025-06-03independentSource ↗
Terminal-BenchActive51.8%
80
2025-06-03independentSource ↗
OpenAI-MRCRActive68.9%
80
2025-06-02independentSource ↗
SimpleQAActive42.6%
90
2025-06-02independentSource ↗
CodeContestsDeprecated48.1%
10
2025-06-02independentSource ↗
ComplexBenchNearing Saturation61.9%
80
2025-06-01independentSource ↗
SWE-bench ProActive45.9%
90
2025-06-01independentSource ↗
LongBench-ProActive67.2%
80
2025-06-01independentSource ↗
ZeroSCROLLSNearing Saturation65.6%
78
2025-05-31independentSource ↗
MultiPL-EDeprecated68.1%
10
2025-05-31independentSource ↗
LiveCodeBenchSaturated60.7%
39
2025-05-30independentSource ↗
AgentIFNearing Saturation51.3%
80
2025-05-30independentSource ↗
MATHSaturated77.9%
25
2025-05-30independentSource ↗
InfiniteBenchNearing Saturation69.7%
65
2025-05-30independentSource ↗
MercuryDeprecated65.1%
10
2025-05-29independentSource ↗
SCROLLSNearing Saturation65.6%
65
2025-05-29independentSource ↗
MMLU-ProNearing Saturation75.4%
75
2025-05-29independentSource ↗
RULERNearing Saturation75.4%
84
2025-05-28independentSource ↗
TabMWPSaturated73.1%
25
2025-05-28independentSource ↗
CyberSecEvalDeprecated66.6%
10
2025-05-28independentSource ↗
SWE-bench VerifiedSaturated57.7%
25
2025-05-27independentSource ↗
MMLU-ReduxSaturated80.4%
25
2025-05-27independentSource ↗
QuALITYNearing Saturation72.2%
78
2025-05-27independentSource ↗
LongBenchNearing Saturation70.5%
65
2025-05-26independentSource ↗
SWE-benchDeprecated50.3%
10
2025-05-26independentSource ↗
MGSMNearing Saturation75.5%
65
2025-05-26independentSource ↗
AGIEvalSaturated69.7%
25
2025-05-25independentSource ↗
QASPERNearing Saturation68.1%
65
2025-05-25independentSource ↗
TruthfulQASaturated72.2%
25
2025-05-25independentSource ↗
GSM-HardDeprecated76.3%
10
2025-05-24independentSource ↗
SuperGLUESaturated82.6%
25
2025-05-24independentSource ↗
CodeXGLUEDeprecated69.6%
23
2025-05-23independentSource ↗
MT-BenchNearing Saturation7.8%
75
2025-05-23independentSource ↗
BIG-Bench HardSaturated79.4%
25
2025-05-23independentSource ↗
WinoGrandeSaturated84.2%
28
2025-05-22independentSource ↗
MBPP+Nearing Saturation75.5%
65
2025-05-22independentSource ↗
IFEvalDeprecated73.9%
25
2025-05-22independentSource ↗
GPQANearing Saturation72.2%
78
2025-05-22independentSource ↗
Needle-in-a-HaystackSaturated96.4%
44
2025-05-21independentSource ↗
HumanEval+Saturated79.9%
25
2025-05-21independentSource ↗
CommonsenseQASaturated85.1%
28
2025-05-21independentSource ↗
MBPPSaturated78.4%
25
2025-05-20independentSource ↗
GSM8KSaturated85.9%
25
2025-05-20independentSource ↗
GLUESaturated87.5%
28
2025-05-20independentSource ↗
MMLUSaturated83.6%
25
2025-05-19independentSource ↗
HumanEvalSaturated84.4%
25
2025-05-19independentSource ↗
ARC-AGISaturated76.1%
25
2025-05-19independentSource ↗
LiveBench (Continuous)Active72.8%
94
2025-02-01independentSource ↗
OpenAI Compatible API ExecutionAPI Ready
from openai import OpenAI

client = OpenAI()
response = client.chat.completions.create(
    model="amazon-nova-2-lite-v1",
    messages=[{"role": "user", "content": "Evaluate multi-step logic problem"}],
    temperature=0.2
)
print(response.choices[0].message.content)

Evaluation & Sourcing Notes

Scores listed for Amazon Nova 2 Lite represent verified evaluations extracted from official research papers, independent evaluation suites (HELM, LMSYS, OpenCompass, LiveBench), and verified audit reports.

All benchmarks marked as saturated or deprecated reflect historical performance where the benchmark no longer provides active discriminative power.