The Open-Source AI Stack
RSS
All models

Models · gpt-3-5

GPT-3.5 Turbo

Proprietary OpenAI · — · Proprietary

Launched alongside ChatGPT on November 30, 2022 as the chat-optimized successor to text-davinci-003 at roughly one-tenth the cost. The 0125 revision settled on a 16K context window and became the default low-cost endpoint through the GPT-4 era.

Cost

$0.50 / Mtok input
$1.50 / Mtok output

OpenAI API · as of 2026-05-21

source ↗

Speed

91.9 tok/sec output
457 ms TTFT

· as of 2026-05-21

source ↗

Architecture

tokens in Embedding vocab not disclosed × N layers Architecture not disclosed (proprietary or undocumented) Output projection tokens out
Schema-generated from data/models.yaml. Every label is auditable against the model's sources.

Specs

Architecture
unknown
Total params
not disclosed
Active params
not disclosed
Context window
not verified
Attention
unknown
Position encoding
unknown
Post-training
rlhf
OSI-approved
no
Data released
no
Training code
not released

Benchmarks

Each score carries the date it was published; we never infer or interpolate missing scores.

General reasoning

MMLU-Pro 46.2 as of 2026-05-21 source ↗
GPQA-Diamond 29.7 as of 2026-05-21 source ↗

Math

MATH 44.1 as of 2026-05-21 source ↗

Available quantizations

None. The weights are not distributed, so there are no public quantizations.

Notable innovations

  • · Chat-optimized API at GPT-3-class cost
  • · Backed the original ChatGPT launch

Sources