No longer supported

This model is retired (released November 2022) and is no longer actively maintained or recommended by OpenAI. Consider using their current flagship model instead.

retiredlegacyllm Proprietary

text-davinci-003

by OpenAI· Released November 2022· Cutoff June 2021

text-davinci-003 is a legacy GPT-3.5 model that was the most capable model in the GPT-3.5 series before GPT-3.5-turbo. It offered improved instruction following, factuality, and consistency compared to its predecessors. It has been superseded by newer models like GPT-3.5-turbo and GPT-4.

Official Site API Docs

Input cost

$0.0200 per 1K tokens

Output cost

$0.0200 per 1K tokens

Context window

4097 tokens

Max output

4096 tokens

Modalities

text

Parameters

175B

License

proprietary

Capabilities

Text GenerationInstruction FollowingCode GenerationZero-shot LearningFew-shot LearningText Completion

Best For

General-purpose text generation and instruction following tasks, especially where high-quality completions are needed.

Strengths

Strong instruction following
Good at zero-shot and few-shot tasks
High-quality text generation
Supports complex prompts

Limitations

Legacy model, no longer recommended
Higher latency compared to newer models
No function calling or JSON mode
No vision capabilities

Use Cases

Content creation

Code generation and explanation

Question answering

Summarization

Translation

Creative writing

Chatbots

Improvements Over Previous Model

Improved instruction following over text-davinci-002
Better factuality and consistency
Reduced toxic outputs
Enhanced ability to handle complex tasks
Improved zero-shot performance

Back to all models

No longer supported

This model is retired (released November 2022) and is no longer actively maintained or recommended by OpenAI. Consider using their current flagship model instead.

retiredlegacyllm Proprietary

text-davinci-003

by OpenAI· Released November 2022· Cutoff June 2021

Official Site API Docs

Input cost

$0.0200 per 1K tokens

Output cost

$0.0200 per 1K tokens

Context window

4097 tokens

Max output

4096 tokens

Modalities

text

Parameters

175B

License

proprietary

Capabilities

Text GenerationInstruction FollowingCode GenerationZero-shot LearningFew-shot LearningText Completion

Best For

General-purpose text generation and instruction following tasks, especially where high-quality completions are needed.

Strengths

Strong instruction following
Good at zero-shot and few-shot tasks
High-quality text generation
Supports complex prompts

Limitations

Legacy model, no longer recommended
Higher latency compared to newer models
No function calling or JSON mode
No vision capabilities

Use Cases

Content creation

Code generation and explanation

Question answering

Summarization

Translation

Creative writing

Chatbots

Improvements Over Previous Model

Improved instruction following over text-davinci-002
Better factuality and consistency
Reduced toxic outputs
Enhanced ability to handle complex tasks
Improved zero-shot performance

Back to all models