GPT-3.5 Turbo: Complete Benchmark Performance Guide

openai 3.5-turbo 2023-03-01 Proprietary

In-depth analysis of GPT-3.5 Turbo performance on MMLU, HumanEval, GSM8K, MATH, BBH benchmarks, including pricing and use cases.

Key Specifications

Vendoropenai
Version3.5-turbo
Release Date2023-03-01
Context Window16385 tokens
Input Modalitiestext
Output Modalitiestext
LicenseProprietary
Documentationhttps://platform.openai.com/docs/models

Benchmark Performance

BenchmarkScoreUnitEvaluated AtNotesSource
MMLU61.1%2023-03-015-shotview
HUMANEVAL51.3pass@12023-03-01—view
GSM8K34.3%2023-03-010-shot CoTview
MATH17.8%2023-03-010-shot CoTview
BBH48.2%2023-03-013-shot CoTview
GPQA22.4%2023-03-010-shotview
IFEVAL57.4%2023-03-01prompt_strictview
ARC86.8%2023-03-01challengeview
MUSR29.2%2023-03-010-shotview
WINOGRANDE76.7%2023-03-010-shotview

Pricing

TierPriceCurrency
Input$0.5 / MtokUSD
Output$1.5 / MtokUSD
Cache Read$0 / MtokUSD
Cache Write$0 / MtokUSD

Source: https://openai.com/api/pricing/ · as of 2023-03-01

Compliance

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

GPT-3.5 Turbo

모델 개요

OpenAI GPT-3.5 Turbo 经济型模型, 16K 上下文, 速度快价格低, 适合对话与通用任务。

핵심 사양

공급업체버전출시일컨텍스트 창입력 모달리티출력 모달리티라이선스
Openai3.5-turbo2023-03-0116KtexttextProprietary

벤치마크 성능

벤치마크점수단위비고
MMLU (Massive Multitask Language Understanding)61.1%5-shot
HumanEval51.3pass@1—
GSM8K (Grade School Math 8K)34.3%0-shot CoT
MATH17.8%0-shot CoT
BBH (BIG-Bench Hard)48.2%3-shot CoT
GPQA22.4%0-shot
IFEval57.4%prompt_strict
ARC86.8%challenge
MUSR29.2%0-shot
WinoGrande76.7%0-shot

가격

입력출력캐시 읽기캐시 쓰기
————

백만 토큰당

강점

  • 可靠的通用模型。

약점

  • 闭源专有模型，不支持自托管。
  • 上下文窗口 16K 偏小。

사용 사례

  • 通用对话与问答

참고문헌