sigoden/aichat · error

No available embedding model

Error message

No available embedding model

What it means

`Rag::create_config` needs an embedding model; when none was selected/known it lists configured embedding models, and if that list is empty there is no valid choice, so it bails with 'No available embedding model'.

Solutions

1. Add an embedding model to the config (e.g. an OpenAI/other provider embedding entry) and verify its API key
2. Select an embedding model interactively so the None branch isn't hit
3. Confirm the model is registered with ModelType::Embedding, not Chat
4. Run `list_models`/model listing command to verify discovery works

Example fix

// before: only a chat model configured
models = [{ name = "gpt-4o", type = "chat" }]
// after
models = [
  { name = "gpt-4o", type = "chat" },
  { name = "text-embedding-3-small", type = "embedding", max_chunk_size = 1000 }
]

let models = list_models(&config.read(), ModelType::Embedding);
if models.is_empty() {
    eprintln!("configure an embedding model before creating a RAG");
}

match Rag::create_config(&config).await {
    Err(e) if e.to_string() == "No available embedding model" => {
        eprintln!("Add an embedding model (and API key) to your config");
    }
    other => other?,
}

When it happens

Trigger: `create_config` reaches the `None` branch of the interactive selection (no pre-selected embedding model) and `list_models(config, ModelType::Embedding)` returns empty.

Common situations: Config has no embedding-capable models defined (e.g. only chat models configured); provider credentials missing so model discovery returns nothing; typo in model type in the config.

Understand the failure class

Background: "is required", "must be set", "missing required field": configuration validation errors across open-source libraries — this error's family across 36 libraries.

Related errors

• No models
• Unknown client
• Unknown model
• Unknown RAG
• No RAG

AI-assisted analysis of sigoden/aichat@82976d349a (2026-09-09). Data as JSON: /api/errors/5c7e0a7a95b07196. Report an issue: GitHub.

Appendix: source

Thrown at src/rag/mod.rs:169

    pub fn create_config(config: &GlobalConfig) -> Result<(Model, usize, usize)> {
        let (embedding_model_id, chunk_size, chunk_overlap) = {
            let config = config.read();
            (
                config.rag_embedding_model.clone(),
                config.rag_chunk_size,
                config.rag_chunk_overlap,
            )
        };
        let embedding_model_id = match embedding_model_id {
            Some(value) => {
                println!("Select embedding model: {value}");
                value
            }
            None => {
                let models = list_models(&config.read(), ModelType::Embedding);
                if models.is_empty() {
                    bail!("No available embedding model");
                }
                select_embedding_model(&models)?
            }
        };
        let embedding_model =
            Model::retrieve_model(&config.read(), &embedding_model_id, ModelType::Embedding)?;

        let chunk_size = match chunk_size {
            Some(value) => {
                println!("Set chunk size: {value}");
                value
            }
            None => set_chunk_size(&embedding_model)?,
        };
        let chunk_overlap = match chunk_overlap {
            Some(value) => {
                println!("Set chunk overlay: {value}");
                value

View on GitHub (pinned to 82976d349a)