Amazon logo

Nova 2 Lite

Multimodal
Amazon

Amazon Nova 2 Lite is a low-latency, cost-effective multimodal reasoning model for everyday workloads.

Key Specifications

Parameters
-
Context
1.0M
Release Date
December 2, 2025
Average Score
68.1%

Timeline

Key dates in the model's history
Announcement
December 2, 2025
Last Update
August 29, 2026
Today
November 20, 2026

Technical Specifications

Parameters
-
Training Tokens
-
Knowledge Cutoff
-
Family
-
Capabilities
MultimodalZeroEval

Pricing & Availability

Input (per 1M tokens)
$0.30
Output (per 1M tokens)
$2.50
Max Input Tokens
1.0M
Max Output Tokens
65.5K
Supported Features
Function CallingStructured OutputCode ExecutionWeb SearchBatch InferenceFine-tuning

Benchmark Results

Model performance metrics across various tests and benchmarks

Programming

Programming skills tests
SWE-Bench Verified
with inference-time scaling and internal agentic scaffolding • Self-reported
64.5%

Reasoning

Logical reasoning and analysis
GPQA
Diamond accuracy • Self-reported
79.6%

Other Tests

Specialized benchmarks
AIME 2025
accuracy • Self-reported
91.0%
BFCL-V4
overall FC accuracy • Self-reported
60.3%
IFBench
prompt loose accuracy • Self-reported
70.8%
LiveCodeBench
v5 (Jul 2024 - Jan 2025 subset) • Self-reported
71.0%
LongCodeBench
1M match accuracy • Self-reported
84.0%
MCP Atlas
pass rate • Self-reported
24.6%
MMLU-Pro
accuracy • Self-reported
80.9%
MMMU-Pro
accuracy • Self-reported
61.8%
Multi-Challenge
accuracy • Self-reported
76.6%
OCRBench_V2
avg accuracy • Self-reported
56.1%
QVHighlights
R1@0.5 • Self-reported
77.2%
RealKIE-FCC
ANLS (Verified) • Self-reported
62.1%
ScreenSpot
point accuracy • Self-reported
83.3%
Tau2 Airline
Verified avg@3 • Self-reported
64.8%
Tau2 Retail
Verified avg@3 • Self-reported
76.5%
Tau2 Telecom
avg@3 (Artificial Analysis) • Self-reported
76.0%
Terminal-Bench
Terminal-Bench 1.0 accuracy • Self-reported
32.5%

License & Metadata

License
proprietary
Announcement Date
December 2, 2025
Last Updated
August 29, 2026

Similar Models

All Models

Recommendations are based on similarity of characteristics: developer organization, multimodality, parameter size, and benchmark performance. Choose a model to compare or go to the full catalog to browse all available AI models.