OpenAI

OpenAIBuilder presets the OpenAI endpoint and auth, and builds on the Responses API wire for completion and streaming. Hand the built client to ChatBuilder::with_model and the rest of chat-rs works the same as any other provider.

Install

Enable the openai feature (crate chat-openai):

chat-rs = { version = "0.5.1", features = ["openai"] }

By default the client reads your key from the OPENAI_API_KEY environment variable. You can also set it explicitly with with_api_key.

Usage

use chat_rs::{ChatBuilder, ChatOutcome, openai::OpenAIBuilder, parts, types::messages};

let client = OpenAIBuilder::new()
    .with_model("gpt-4o-mini")
    .build();

let mut chat = ChatBuilder::new().with_model(client).build();

let mut messages = messages::from_user(parts!["Hey there!"]);

if let ChatOutcome::Complete(res) = chat.complete(&mut messages).await.map_err(|e| e.err)? {
    println!("{:#?}", res.content);
}

Configuration

The builder methods cover the OpenAI-specific surface:

• with_model(impl Into<String>) selects the model and moves the builder into its ready-to-build state.
• with_api_key(String) supplies the key directly instead of reading OPENAI_API_KEY.
• with_custom_url(String) points the client at a custom OpenAI-compatible endpoint.
• with_store(bool) controls whether responses are stored server-side.
• without_previous_response_id() stops the client from chaining requests via the previous response id, so each call sends full context instead.
• with_reasoning_effort(&str) sets reasoning effort to "low", "medium", or "high" on models that support it.
• with_description, with_metadata, and with_transport attach metadata or swap in a custom transport.

Native tools (see native tools) are added on the builder:

• with_web_search(context_size, user_location) enables OpenAI's web search.
• with_image_generation(tool) enables image generation. This is Beta; the generated bytes arrive back as image file parts.

use chat_rs::openai::{ImageGenerationTool, ImageQuality, ImageSize, OpenAIBuilder};

let client = OpenAIBuilder::new()
    .with_model("gpt-4.1")
    .with_image_generation(ImageGenerationTool {
        size: Some(ImageSize::Square),
        quality: Some(ImageQuality::Auto),
        ..Default::default()
    })
    .build();

For embeddings, switch the builder into embedding mode with with_embeddings():

let client = OpenAIBuilder::new()
    .with_model("text-embedding-3-small")
    .with_embeddings()
    .build();

let mut chat = ChatBuilder::new().with_model(client).with_embeddings().build();
let response = chat.embed(&mut messages).await.map_err(|e| e.err)?;
println!("{}", response.embeddings.dimension);

Capabilities

• Completion and streaming are both supported.
• Embeddings and structured output are supported.
• Multimodal input works by attaching file parts to a user message.
• Native tools: web search and image generation.
• Input streaming lets you push text while the model is producing output, built on the Responses API wire. Enable it with ChatBuilder::with_input_stream.