Providers

Submodel

AISDK provides first-class support for Submodel with fully typed model APIs. Model capabilities are enforced at compile time using Rust's type system. This prevents model capability mismatches and guarantees the selected model is valid for the task (e.g. tool calling).

Installation

Enable the Submodel provider feature:

cargo add aisdk --features submodel

This installs AISDK with the Submodel provider enabled. Once you have enabled the Submodel provider, you can use all aisdk features with it.

Create a Provider Instance

To create a provider instance, call Submodel::model_name(), where model_name is the Submodel model you want to use. Model names are exposed as snake-case methods.

use aisdk::providers::Submodel;

let submodel = Submodel::qwen_qwen3_235b_a22b_instruct_2507();

This initializes the provider with:

  • Model: "Qwen/Qwen3-235B-A22B-Instruct-2507"
  • API key from environment (if set with SUBMODEL_INSTAGEN_ACCESS_KEY)
  • Submodel's default base URL (https://llm.submodel.ai/v1)

Basic Text Generation

Example using LanguageModelRequest for text generation.

use aisdk::{
    core::LanguageModelRequest,
    providers::Submodel,
};

#[tokio::main]
async fn main() -> Result<(), Box<dyn std::error::Error>> {

    let submodel = Submodel::qwen_qwen3_235b_a22b_instruct_2507();

    let response = LanguageModelRequest::builder()
        .model(submodel)
        .prompt("Write a short poem about Rust.")
        .build()
        .generate_text()
        .await?;

    println!("Response text: {:?}", response.text());
    Ok(())
}

Provider Settings

You can customize provider configuration using Submodel::builder()

API Key

let submodel = Submodel::<QwenQwen3235bA22bInstruct2507>::builder()
    .api_key("your-api-key")
    .build()?;

If not specified, AISDK uses the SUBMODEL_INSTAGEN_ACCESS_KEY environment variable.

Base URL

Useful when routing through a proxy, gateway, or self-hosted compatible endpoint.

let submodel = Submodel::<QwenQwen3235bA22bInstruct2507>::builder()
    .base_url("https://llm.submodel.ai/v1")
    .build()?;

Path (Full URL Override)

Use .path(...) to override the full request URL instead of only the base URL.

let submodel = Submodel::<QwenQwen3235bA22bInstruct2507>::builder()
    .path("https://full-url.example/v1/chat/completions")
    .build()?;

Provider Name

For logging, analytics, and observability.

let submodel = Submodel::<QwenQwen3235bA22bInstruct2507>::builder()
    .provider_name("Submodel")
    .build()?;

Full Custom Configuration Example

let submodel = Submodel::<QwenQwen3235bA22bInstruct2507>::builder()
    .api_key("your-api-key")
    .base_url("https://llm.submodel.ai/v1")
    .path("https://full-url.example/v1/chat/completions")
    .provider_name("Submodel")
    .build()?;

Dynamic Model Selection

For runtime model selection (e.g., loading models from config files), use DynamicModel:

Using model_name() Method with Default Settings

use aisdk::providers::Submodel;

// Specify model as a string at runtime
let submodel = Submodel::model_name("Qwen/Qwen3-235B-A22B-Instruct-2507");

Using Builder Pattern with Custom Settings

use aisdk::{
    core::DynamicModel,
    providers::Submodel,
};

let submodel = Submodel::<DynamicModel>::builder()
	.model_name("Qwen/Qwen3-235B-A22B-Instruct-2507")
	.api_key("your-api-key")
	.base_url("https://llm.submodel.ai/v1")
	.path("https://full-url.example/v1/chat/completions")
	.provider_name("Submodel")
	.build()?;

Warning: When using DynamicModel, model capabilities are not validated at compile time. This means there's no guarantee the model supports requested features (e.g., tool calls, structured output). For compile-time safety, use the typed methods like Submodel::qwen_qwen3_235b_a22b_instruct_2507().

Next Steps