Claude 3.5 Sonnet (2024-10-22): Complete Benchmark Performance Guide

anthropic 3.5-sonnet-20241022 2024-10-22 Proprietary

In-depth analysis of Claude 3.5 Sonnet (2024-10-22) performance on MMLU, HumanEval, GSM8K, MATH, BBH benchmarks, including pricing and use cases.

Key Specifications

Vendoranthropic
Version3.5-sonnet-20241022
Release Date2024-10-22
Context Window200000 tokens
Input Modalitiestext, image
Output Modalitiestext
LicenseProprietary
Documentationhttps://docs.anthropic.com/claude/docs

Benchmark Performance

Benchmark Score Unit Evaluated At Notes Source
MMLU 87.2 % 2024-10-22 5-shot view
HUMANEVAL 86.1 pass@1 2024-10-22 — view
GSM8K 94 % 2024-10-22 0-shot CoT view
MATH 75.1 % 2024-10-22 0-shot CoT view
BBH 84.2 % 2024-10-22 3-shot CoT view
GPQA 50.9 % 2024-10-22 0-shot view
IFEVAL 81.6 % 2024-10-22 prompt_strict view
ARC 96.8 % 2024-10-22 challenge view
MUSR 71.8 % 2024-10-22 0-shot view
WINOGRANDE 86.9 % 2024-10-22 0-shot view

Pricing

Tier Price Currency
Input$3 / MtokUSD
Output$15 / MtokUSD
Cache Read$0 / MtokUSD
Cache Write$0 / MtokUSD

Source: https://www.anthropic.com/pricing · as of 2024-10-22

Compliance

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

Claude 3.5 Sonnet (2024-10-22)

Model Overview

Anthropic Claude 3.5 Sonnet 2024-10-22 版本, 引入计算机使用能力, 编码与推理性能再次提升。

Core Specifications

Vendor Version Release Date Context Window Input Modalities Output Modalities License
Anthropic 3.5-sonnet-20241022 2024-10-22 200K text, image text Proprietary

Benchmark Performance

Benchmark Score Unit Notes
MMLU (Massive Multitask Language Understanding) 87.2 % 5-shot
HumanEval 86.1 pass@1 —
GSM8K (Grade School Math 8K) 94.0 % 0-shot CoT
MATH 75.1 % 0-shot CoT
BBH (BIG-Bench Hard) 84.2 % 3-shot CoT
GPQA 50.9 % 0-shot
IFEval 81.6 % prompt_strict
ARC 96.8 % challenge
MUSR 71.8 % 0-shot
WinoGrande 86.9 % 0-shot

Pricing

Input Output Cache Read Cache Write
— — — —

per million tokens

Strengths

  • MMLU score 87.2, strong knowledge reasoning.
  • HumanEval 86.1, excellent code generation.
  • GSM8K 94.0, robust math reasoning.
  • Context window 200K.

Weaknesses

  • Proprietary, not self-hostable.

Use Cases

  • Code generation and debugging
  • Long document summarization
  • Agent workflows and tool use

References