Understand and compare
GPT-4 Turbo
vs.
GPT-3.5 Turbo
Try
Podial
Turn your documents into engaging podcast discussions.
Overview
GPT-4 Turbo was released
11 months after
GPT-3.5 Turbo.
GPT-4 Turbo
|
GPT-3.5 Turbo
|
|
|---|---|---|
|
Provider
The entity that provides this model.
|
OpenAI
|
OpenAI
|
|
Input Context Window
The number of tokens supported by the input context window.
|
128K
tokens
|
4,096
tokens
|
|
Maximum Output Tokens
The number of tokens that can be generated by the model in a single request.
|
4,096
tokens
|
4,096
tokens
|
|
Release Date
When the model was first released.
|
2023-11-06
|
2022-11-28
|
|
Knowledge Cutoff
Limit on the knowledge base used by the model.
|
December 2023
|
September 2021
|
|
Open Source
|
|
|
|
API Providers
The providers that offer this model. (This is not an exhaustive list.)
|
|
|
Pricing
GPT-4 Turbo is
roughly 20x more expensive compared
to GPT-3.5 Turbo for input and output tokens.
|
GPT-4 Turbo
|
GPT-3.5 Turbo
|
|
|---|---|---|
|
Input
Cost of input data provided to the model.
|
$10.00
per million tokens
|
$0.50
per million tokens
|
|
Output
Cost of output tokens generated by the model.
|
$30.00
per million tokens
|
$1.50
per million tokens
|
Benchmarks
Compare relevant benchmarks between GPT-4 Turbo
and GPT-3.5 Turbo.
|
GPT-4 Turbo
|
GPT-3.5 Turbo
|
|
|---|---|---|
|
MMLU
Evaluating LLM knowledge acquisition in zero-shot and few-shot settings.
|
Benchmark not available.
|
70.0
(5-shot)
|
|
MMMU
A wide ranging multi-discipline and multimodal benchmark.
|
Benchmark not available.
|
Benchmark not available.
|
|
HellaSwag
A challenging sentence completion benchmark.
|
Benchmark not available.
|
85.5
(10-shot)
|
|
GSM8K
Grade-school math problems benchmark.
|
Benchmark not available.
|
Benchmark not available.
|
|
HumanEval
A benchmark to measure functional correctness for synthesizing programs from docstrings.
|
Benchmark not available.
|
Benchmark not available.
|
|
MATH
Benchmark performance on Math problems ranging across 5 levels of difficulty and 7 sub-disciplines.
|
Benchmark not available.
|
43.1
(0-shot)
|
GPT-4 Turbo, developed by OpenAI, features a large context window of 128,000 tokens. The model costs 1.0 cent per thousand tokens for input and 3.0 cents per thousand tokens for output. It is set to be released on November 6, 2023.
GPT-3.5 Turbo, developed by OpenAI, features a context window of 4096 tokens. It is priced at 0.05 cents per thousand tokens for input and 0.15 cents per thousand tokens for output. The model was released on November 28, 2022, and has achieved high scores in benchmarks like HellaSwag (85.5 in a 10-shot scenario) and MMLU (70.0 in a 5-shot scenario).
Mistral Large 2
Llama 3 70B Instruct
Claude 3 Sonnet
Measure & Improve LLM
Product Performance.
Get Started
© 2023-2024 Context