Claude 3 Haiku (2024-03-07): Complete Benchmark Performance Guide

anthropic 3-haiku-20240307 2024-03-07 Proprietary

In-depth analysis of Claude 3 Haiku (2024-03-07) performance on MMLU, HumanEval, GSM8K, MATH, BBH benchmarks, including pricing and use cases.

Key Specifications

Vendoranthropic
Version3-haiku-20240307
Release Date2024-03-07
Context Window200000 tokens
Input Modalitiestext, image
Output Modalitiestext
LicenseProprietary
Documentationhttps://docs.anthropic.com/claude/docs

Benchmark Performance

BenchmarkScoreUnitEvaluated AtNotesSource
MMLU77.1%2024-03-075-shotview
HUMANEVAL74.8pass@12024-03-07—view
GSM8K79.5%2024-03-070-shot CoTview
MATH44.7%2024-03-070-shot CoTview
BBH72.2%2024-03-073-shot CoTview
GPQA39.6%2024-03-070-shotview
IFEVAL75.3%2024-03-07prompt_strictview
ARC91.6%2024-03-07challengeview
MUSR51%2024-03-070-shotview
WINOGRANDE79.5%2024-03-070-shotview

Pricing

TierPriceCurrency
Input$0.25 / MtokUSD
Output$1.25 / MtokUSD
Cache Read$0 / MtokUSD
Cache Write$0 / MtokUSD

Source: https://www.anthropic.com/pricing · as of 2024-03-07

Compliance

  • Data Residency: US
  • SOC2: ✓
  • HIPAA: ✗
  • GDPR: ✓
  • ISO 27001: ✗

Claude 3 Haiku (2024-03-07)

Aperçu du modèle

Anthropic Claude 3 Haiku 2024-03-07 快照版本, 200K 上下文, 最快响应速度的经济型模型。

Spécifications principales

FournisseurVersionDate de sortieFenêtre de contexteModalités d’entréeModalités de sortieLicence
Anthropic3-haiku-202403072024-03-07200Ktext, imagetextProprietary

Performance aux benchmarks

BenchmarkScoreUnitéNotes
MMLU (Massive Multitask Language Understanding)77.1%5-shot
HumanEval74.8pass@1—
GSM8K (Grade School Math 8K)79.5%0-shot CoT
MATH44.7%0-shot CoT
BBH (BIG-Bench Hard)72.2%3-shot CoT
GPQA39.6%0-shot
IFEval75.3%prompt_strict
ARC91.6%challenge
MUSR51.0%0-shot
WinoGrande79.5%0-shot

Tarification

EntréeSortieLecture cacheÉcriture cache
————

par million de jetons

Forces

  • 可靠的通用模型。

Faiblesses

  • 闭源专有模型，不支持自托管。

Cas d’usage

  • 代码生成与调试

Références