Pass IndexThe State of AISign in

GPT-3.5 Turbo (older v0613)

GPT-3.5 Turbo is OpenAI's fastest model.

text → text · made by OpenAI

$1$2per Mtok in / out
Nous Research · 2 sellers

Sold by 2 ways

SellerLaneRate
Nous Research aggregatorstandard$1per Mtok in$2per Mtok outinference-api.nousresearch.com · read 2026-09-12
OpenRouter aggregatorstandard$1per Mtok in$2per Mtok outopenrouter.ai · read 2026-08-24

Measured 7 standings

PlaceBoardMetricScore
3rdof 6Common sense qa 2 — Epoch AIScore57
17thof 137Wino grande — Epoch AIAccuracy81.6
18thof 88Bbh — Epoch AIAverage0.616
21stof 136Bool q — Epoch AIScore87
40thof 170Gsm8k — Epoch AIEM0.578
77thof 217Mmlu — Epoch AIEM0.689
607thof 656Epoch capabilities index — Epoch AIECI Score113.17

About

GPT-3.5 Turbo (older v0613) — a text model from OpenAI, sold by 2 companies from $1 in and $2 out per million tokens, placed 17th of 137 on Wino grande — Epoch AI.

It takes text and returns text, with a context window of 4,095 tokens. It was published in January 2024, trained on material up to September 2021. Its sellers say it can call a tool. The catalogue files it under chat. Two companies sell it. The cheapest is $1 in and $2 out per million tokens at Nous Research. It has been measured on 7 boards, and stands best at 17th of 137 on Wino grande — Epoch AI.

Every current figure

Maker
OpenAI
Register
model
Takes
text
Returns
text
Context
4,095 tokens
Longest answer
3,685 tokens
Published
January 2024
Knowledge to
September 2021
Licence
not read
Sellers
2
Maker's own price
not read
Price
$1 in and $2 out per million tokens — Nous Research
Boards
7
Best place
17th of 137 — Wino grande — Epoch AI

Known as 1 name

openai/gpt-3.5-turbo-0613