Mintplex-Labs/anything-llm · critical · Error

OMLX must have a valid model set.

Error message

OMLX must have a valid model set.

What it means

OMLXProvider's constructor throws synchronously 'OMLX must have a valid model set.' when neither config.model nor process.env.OMLX_LLM_MODEL_PREF resolves to a truthy value. The OpenAI-compatible client for the OMLX (OpenMLX) service is constructed on the next lines and requires a concrete model for every call, so this is a fail-fast configuration guard at instantiation time — no network request has happened when you see it.

Solutions

1. Set OMLX_LLM_MODEL_PREF to a valid OMLX model id in the server environment and restart
2. Or pass the model explicitly: new OMLXProvider({ model: "..." }) / choose the model in the agent provider settings
3. Verify OMLX_LLM_BASE_PATH and OMLX_LLM_API_KEY are also set so the provider works once constructed

Example fix

# before
# OMLX_LLM_MODEL_PREF unset -> constructor throws
OMLX_LLM_BASE_PATH=https://api.omlx.ml/v1

# after
OMLX_LLM_MODEL_PREF=meta-llama/Llama-3.1-8B-Instruct
OMLX_LLM_BASE_PATH=https://api.omlx.ml/v1
OMLX_LLM_API_KEY=<key>

// Guard before construction
const model = agentConfig?.model || process.env.OMLX_LLM_MODEL_PREF;
if (!model) throw new Error("Set OMLX_LLM_MODEL_PREF or pass { model } before using the OMLX agent");
const provider = new OMLXProvider({ model });

const hasOmlxModel = (config) => Boolean(config?.model || process.env.OMLX_LLM_MODEL_PREF);

try {
  provider = new OMLXProvider(config);
} catch (error) {
  if (/valid model set/i.test(error.message)) {
    return respond("Set OMLX_LLM_MODEL_PREF to a model your OMLX endpoint serves, then retry.");
  }
  throw error;
}

When it happens

Trigger: new OMLXProvider() (or with model unset) while OMLX_LLM_MODEL_PREF is missing/empty in the environment — typically during agent initialization before the first chat turn.

Common situations: OMLX provider selected in settings but the model env never populated; hand-edited .env dropped the variable; deployment platform secrets not injected into the container; workspace agent saved without a model and no global default.

Related errors

• LMStudio must have a valid model set.
• GenericOpenAI must have a valid base path to use for the…
• LiteLLM must have a valid base path to use for the api.
• No embedding base path was set.
• No embedding base path was set.

AI-assisted analysis of Mintplex-Labs/anything-llm@3aec848f28 (2026-08-18). Data as JSON: /api/errors/c5656baf6e995de2. Report an issue: GitHub.

Appendix: source

Thrown at server/utils/agents/aibitat/providers/omlx.js:24

const { RetryError } = require("../error.js");
const { parseOMLXBasePath } = require("../../../AiProviders/omlx/index.js");

/**
 * The agent provider for OMLX.
 * OMLX is OpenAI-compatible and supports native tool calling, falling back
 * to the UnTooled prompt-based approach when disabled via env.
 */
class OMLXProvider extends InheritMultiple([Provider, UnTooled]) {
  model;

  /**
   * @param {{model?: string}} config
   */
  constructor(config = {}) {
    super();
    this.providerTag = "omlx";
    const model = config?.model || process.env.OMLX_LLM_MODEL_PREF;
    if (!model) throw new Error("OMLX must have a valid model set.");

    const client = new OpenAI({
      baseURL: parseOMLXBasePath(process.env.OMLX_LLM_BASE_PATH),
      apiKey: process.env.OMLX_LLM_API_KEY || null,
    });

    this._client = client;
    this.model = model;
    this.verbose = true;
  }

  get client() {
    return this._client;
  }

  get supportsAgentStreaming() {
    return true;
  }

View on GitHub (pinned to 3aec848f28)