GPT-3.5 / ChatGPT

OpenAI · November 2022

deprecatedCloseddecoder onlytextAPI Available
Parameters175B
Context Window4K tokens
VariantsGPT-3.5, GPT-3.5 Turbo, GPT-3.5 Turbo 16K
Sunset DateJune 2025

Why It Matters

Brought AI to the mainstream. Reached 100 million users in 2 months — the fastest-growing consumer application in history at the time. Sparked a global AI arms race among tech companies.

Description

The model behind ChatGPT — the chatbot that brought AI to the mainstream. Built on GPT-3's foundation with RLHF (reinforcement learning from human feedback) to make it conversational, helpful, and safe. Its free chat interface made powerful AI accessible to anyone with a web browser for the first time.

Notable Milestones

  • Reached 100 million users faster than any consumer app in history
  • Triggered Google to declare a 'code red' and accelerate AI efforts
  • Used by students, writers, and professionals worldwide for writing, brainstorming, and learning
  • Inspired every major tech company to launch competing AI chatbots

Benchmark Scores

MMLUMassive Multitask Language Understanding — 57 subjects
70.0%
HumanEvalCode generation pass@1 — Python problems
48.1%

Key Innovations

RLHF
RLHFReinforcement Learning from Human Feedback — training models to align with human preferences by having humans rank outputs.
Instruction Tuning
Instruction TuningFine-tuning a model on instruction-response pairs so it follows user commands more reliably.

Family Tree

Lineage

Successors (1)

Related Research (2)

2017 · OpenAI / DeepMind

Pioneered the RLHF paradigm — training a reward model from human preferences, then using it to fine-tune policies via reinforcement learning.

InstructGPTAlignment
2022 · OpenAI

Applied RLHF to GPT-3: supervised fine-tuning → reward modeling → PPO optimization. Made models safer, more helpful, and more aligned.

External Links