Text Generation ModelDeprecated

GPT-3.5 Instruct

GPT-3.5 Instruct is an instruction-tuned completion model from OpenAI with a 4,096 token context window.

Start Building with GPT-3.5 Instruct View All Models

Publisher

OpenAI

TypeInstruct

Context Window4,096 tokens

Replaced byGPT-5 mini

Try GPT-3.5 Instruct →

About GPT-3.5 Instruct

Instruction-following completions from OpenAI

GPT-3.5 Instruct (gpt-3.5-turbo-instruct) is a variant of OpenAI's GPT-3.5 series fine-tuned specifically to follow user instructions more precisely. Unlike the chat-oriented GPT-3.5 Turbo, this model uses the older completions API format rather than the chat completions format, making it suitable for workflows that expect a single prompt-in, completion-out interaction pattern. It was designed to better understand and adhere to explicit directives in a prompt, improving reliability for task-specific applications.

GPT-3.5 Instruct is best suited for applications that require direct instruction-following in a traditional text completion format, such as structured text generation, templated content, or legacy integrations built around the completions endpoint. The model has a context window of 4,096 tokens and a maximum response size of 2,000 tokens. OpenAI has marked this model as deprecated, so developers building new applications are encouraged to evaluate current alternatives in the OpenAI model lineup.

Capabilities

What GPT-3.5 Instruct supports

Instruction Following

Responds to explicit user directives in a prompt-completion format, tuned to adhere closely to task-specific instructions rather than open-ended conversation.

Text Completion

Uses the legacy completions API endpoint (not chat completions), accepting a single prompt string and returning a generated text continuation.

Structured Output

Can generate structured text such as formatted lists, templates, or fill-in-the-blank content when given explicit formatting instructions in the prompt.

Context Handling

Supports a 4,096 token context window, with a maximum response size of 2,000 tokens per request.

Ready to build with GPT-3.5 Instruct?

Get Started Free

FAQ

Common questions about GPT-3.5 Instruct

What is the context window for GPT-3.5 Instruct?

GPT-3.5 Instruct has a context window of 4,096 tokens, which includes both the input prompt and the generated response. The maximum response size is 2,000 tokens.

How does GPT-3.5 Instruct differ from GPT-3.5 Turbo?

GPT-3.5 Instruct uses the legacy completions API format (a single prompt string returning a completion), whereas GPT-3.5 Turbo uses the chat completions API format with a structured list of messages. GPT-3.5 Instruct was also fine-tuned specifically for instruction-following rather than multi-turn conversation.

Is GPT-3.5 Instruct still available?

GPT-3.5 Instruct has a status of deprecated according to the model metadata. Developers building new applications should check OpenAI's current model offerings for actively supported alternatives.

What is the knowledge cutoff for GPT-3.5 Instruct?

A specific knowledge cutoff date is not provided in the available metadata for this model. OpenAI's GPT-3.5 series generally has a training data cutoff of early 2022, but you should verify the exact date in OpenAI's official documentation.

Does GPT-3.5 Instruct support image or video inputs?

No. Based on the model metadata, GPT-3.5 Instruct does not support image or video analysis. It accepts text-only inputs via the completions endpoint.

Resources