Nova Pro

AmazonAmazonProprietary

描述

Amazon Nova Pro is a highly-capable multimodal model with state-of-the-art performance across text, image, and video understanding. It excels at core capabilities like language understanding, mathematical reasoning, and multimodal tasks while offering industry-leading speed and cost efficiency.

发布日期

2024-12-03

参数规模

—

上下文长度

300K

支持模态

image, pdf, text, video

能力雷达图

general

coding

reasoning

science

agents

multimodal

排行榜排名

领域	#排名	分数	来源
代码能力榜	516	19.0	AA
通用能力榜	498	28.0	AA
多模态榜	127	35.0	LS
科学能力	523	25.0	AA

基准测试分数 (LLM Stats)

(LLM Stats (zeroeval))

Chat

IFEvalGoogle Research (2023)

92.1%自报

Finance

FinQA

77.2%自报

General

MMLU

85.9%自报

BFCL

68.4%自报

Language

Translation en→Set1 COMET22

89.1%自报

Translation Set1→en COMET22

89.0%自报

Translation Set1→en spBleu

44.4%自报

Translation en→Set1 spBleu

43.4%自报

Math

GSM8k

94.8%自报

MATH

76.6%自报

Multimodal

GroundUI-1K

81.4%自报

VATEX

77.8%自报

MM-Mind2Web

63.7%自报

MMMU

61.7%自报

Reasoning

ARC-C

94.8%自报

ChartQAMasry et al. (2022)

89.2%自报

HumanEvalOpenAI (2021)

89.0%自报

BBH

86.9%自报

DROP

85.4%自报

CRAG

50.3%自报

GPQANYU + Cohere + Anthropic (2023)

46.9%自报

Summarization

SQuALITY

19.8%自报

Vision

DocVQADocVQA (2020)

93.5%自报

TextVQA

81.5%自报

VisualWebBench

79.7%自报

EgoSchema

72.1%自报

LVBench

41.6%自报

AA 评测指数

(Artificial Analysis)

Math 500(OpenAI (2024), subset of Hendrycks et al. MATH (2021))

78.6

Mmlu Pro(TIGER-Lab (Univ. of Waterloo, Toronto, CMU, 2024))

69.1

Gpqa(NYU + Cohere + Anthropic (2023))

49.9

Ifbench(Google Research (2023))

38.1

Livecodebench(UC Berkeley + MIT + Cornell (2024))

23.3

Lcr(Artificial Analysis)

21.0

Tau2(Sierra + U Toronto + Vector Institute (2025))

14.0

Aime(MAA (Mathematical Association of America))

10.7

Aime 25(MAA (Mathematical Association of America))

7.0

Intelligence Index(Artificial Analysis)

7.0

Math Index(Artificial Analysis)

7.0

Terminalbench Hard(Stanford × Laude Institute (2026))

6.1

Hle(Center for AI Safety + Scale AI (2025))

3.2

LLM Stats 分类评分

(LLM Stats (zeroeval))

Chat

Image To Text

Instruction Following

Legal

Structured Output

Code

Math

Reasoning

Language

Multimodal

Finance

General

Healthcare

Tool Calling

Vision

Frontend Development

Agents

Economics

Physics

Biology

Chemistry

Long Context

Summarization

定价

输入价格$0.8 / 1M tokens

输出价格$3.2 / 1M tokens

混合价格(3:1)$1.4 / 1M tokens

缓存读取价格$0.2 / 1M tokens

缓存写入价格$0.8 / 1M tokens

速度

Tokens/秒0.0

首Token延迟0.00s

首回答延迟0.00s

供应商价格排行

2 个供应商

最便宜: Amazon最贵: Amazon Bedrock

供应商输入输出

1Amazon主要

$0.8

$3.2

2Amazon Bedrock

$0.8

$3.2

比较该模型在不同 API 供应商之间的定价。

外部链接

LLM Stats Artificial Analysis