GPT升级记录
Models
Overview
The OpenAI API is powered by a diverse set of models with different capabilities and price points. You can also make limited customizations to our original base models for your specific use case with fine-tuning.
MODELS | DESCRIPTION |
GPT-4 Limited beta | A set of models that improve on GPT-3.5 and can understand as well as generate natural language or code |
A set of models that improve on GPT-3 and can understand as well as generate natural language or code | |
DALL·E Beta | A model that can generate and edit images given a natural language prompt |
Beta | A model that can convert audio into text |
A set of models that can convert text into a numerical form | |
Codex Limited beta | A set of models that can understand and generate code, including translating natural language to code |
A fine-tuned model that can detect whether text may be sensitive or unsafe | |
GPT-3 | A set of models that can understand and generate natural language |
We have also published open source models including Point-E, Whisper, Jukebox, and CLIP.
Visit our model index for researchers to learn more about which models have been featured in our research papers and the differences between model series like InstructGPT and GPT-3.5.
GPT-4 is a large multimodal model (accepting text inputs and emitting text outputs today, with image inputs coming in the future) that can solve difficult problems with greater accuracy than any of our previous models, thanks to its broader general knowledge and advanced reasoning capabilities. Like , GPT-4 is optimized for chat but works well for traditional completions tasks.gpt-3.5-turbo
GPT-4 is currently in a limited beta and only accessible to those who have been granted access. Please join the waitlist to get access when capacity is available.
LATEST MODEL | DESCRIPTION | MAX TOKENS | TRAINING DATA |
gpt-4 | More capable than any GPT-3.5 model, able to do more complex tasks, and optimized for chat. Will be updated with our latest model iteration. | 8,192 tokens | Up to Sep 2021 |
gpt-4-0314 | Snapshot of from March 14th 2023. Unlike , this model will not receive updates, and will only be supported for a three month period ending on June 14th 2023. | 8,192 tokens | Up to Sep 2021 |
gpt-4-32k | Same capabilities as the base mode but with 4x the context length. Will be updated with our latest model iteration. | 32,768 tokens | Up to Sep 2021 |
gpt-4-32k-0314 | Snapshot of from March 14th 2023. Unlike , this model will not receive updates, and will only be supported for a three month period ending on June 14th 2023. | 32,768 tokens | Up to Sep 2021 |
For many basic tasks, the difference between GPT-4 and GPT-3.5 models is not significant. However, in more complex reasoning situations, GPT-4 is much more capable than any of our previous models.
GPT-3.5
GPT-3.5 models can understand and generate natural language or code. Our most capable and cost effective model in the GPT-3.5 family is which has been optimized for chat but works well for traditional completions tasks as well.gpt-3.5-turbo
LATEST MODEL | DESCRIPTION | MAX TOKENS | TRAINING DATA |
gpt-3.5-turbo | Most capable GPT-3.5 model and optimized for chat at 1/10th the cost of . Will be updated with our latest model iteration. | 4,096 tokens | Up to Sep 2021 |
gpt-3.5-turbo-0301 | Snapshot of from March 1st 2023. Unlike , this model will not receive updates, and will only be supported for a three month period ending on June 1st 2023. | 4,096 tokens | Up to Sep 2021 |
text-davinci-003 | Can do any language task with better quality, longer output, and consistent instruction-following than the curie, babbage, or ada models. Also supports inserting completions within text. | 4,097 tokens | Up to Jun 2021 |
text-davinci-002 | Similar capabilities to but trained with supervised fine-tuning instead of reinforcement learning | 4,097 tokens | Up to Jun 2021 |
code-davinci-002 | Optimized for code-completion tasks | 8,001 tokens | Up to Jun 2021 |
We recommend using over the other GPT-3.5 models because of its lower cost.gpt-3.5-turbo
OpenAI models are non-deterministic, meaning that identical inputs can yield different outputs. Setting temperature to 0 will make the outputs mostly deterministic, but a small amount of variability may remain.
Feature-specific models
While the new model is optimized for chat, it works very well for traditional completion tasks. The original GPT-3.5 models are optimized for text completion.gpt-3.5-turbo
Our endpoints for creating embeddings and editing text use their own sets of specialized models.
Finding the right model
Experimenting with is a great way to find out what the API is capable of doing. After you have an idea of what you want to accomplish, you can stay with or another model and try to optimize around its capabilities.gpt-3.5-turbogpt-3.5-turbo
You can use the GPT comparison tool that lets you run different models side-by-side to compare outputs, settings, and response times and then download the data into an Excel spreadsheet.
DALL·E is a AI system that can create realistic images and art from a description in natural language. We currently support the ability, given a prommpt, to create a new image with a certain size, edit an existing image, or create variations of a user provided image.
The current DALL·E model available through our API is the 2nd iteration of DALL·E with more realistic, accurate, and 4x greater resolution images than the original model. You can try it through the our Labs interface or via the API.
Whisper is a general-purpose speech recognition model. It is trained on a large dataset of diverse audio and is also a multi-task model that can perform multilingual speech recognition as well as speech translation and language identification. The Whisper v2-large model is currently available through our API with the model name.whisper-1
Currently, there is no difference between the open source version of Whisper and the version available through our API. However, through our API, we offer an optimized inference process which makes running Whisper through our API much faster than doing it through other means. For more technical details on Whisper, you can read the paper.
Embeddings
Embeddings are a numerical representation of text that can be used to measure the relateness between two pieces of text. Our second generation embedding model, is a designed to replace the previous 16 first-generation embedding models at a fraction of the cost. Embeddings are useful for search, clustering, recommendations, anomaly detection, and classification tasks. You can read more about our latest embedding model in the announcement blog post. text-embedding-ada-002
The Codex models are descendants of our GPT-3 models that can understand and generate code. Their training data contains both natural language and billions of lines of public code from GitHub. Learn more.
They’re most capable in Python and proficient in over a dozen languages including JavaScript, Go, Perl, PHP, Ruby, Swift, TypeScript, SQL, and even Shell.
We currently offer two Codex models:
LATEST MODEL | DESCRIPTION | MAX TOKENS | TRAINING DATA |
code-davinci-002 | Most capable Codex model. Particularly good at translating natural language to code. In addition to completing code, also supports inserting completions within code. | 8,001 tokens | Up to Jun 2021 |
code-cushman-001 | Almost as capable as Davinci Codex, but slightly faster. This speed advantage may make it preferable for real-time applications. | Up to 2,048 tokens |
For more, visit our guide on working with Codex.
