Mintplex-Labs/anything-llm · error

No LocalAI Base Path was set.

Error message

No LocalAI Base Path was set.

What it means

LocalAiLLM's constructor throws when LOCAL_AI_BASE_PATH is unset. LocalAI is a self-hosted, OpenAI-compatible drop-in server; this URL is the OpenAI SDK baseURL (LOCAL_AI_API_KEY is optional, defaulting to null). The check is eager at provider construction, and — unlike some siblings — no default model throw follows: LOCAL_AI_MODEL_PREF is read with || and may stay undefined, deferring failure to the HTTP call.

Solutions

1. Confirm LocalAI answers: curl http://127.0.0.1:8080/v1/models.
2. Set LOCAL_AI_BASE_PATH to that base (e.g., http://127.0.0.1:8080/v1) in the backend .env.
3. Docker: use http://host.docker.internal:8080/v1 or the compose service name.
4. Also set LOCAL_AI_MODEL_PREF (required for requests even though the constructor won't complain) and LOCAL_AI_API_KEY if enabled; then restart.

Example fix

# before (.env)
LOCAL_AI_MODEL_PREF=luna-ai-llama2

# after (.env)
LOCAL_AI_BASE_PATH=http://127.0.0.1:8080/v1
LOCAL_AI_MODEL_PREF=luna-ai-llama2

if (!process.env.LOCAL_AI_BASE_PATH) throw new Error("LOCAL_AI_BASE_PATH missing");
try { new URL(process.env.LOCAL_AI_BASE_PATH); } catch { throw new Error("LOCAL_AI_BASE_PATH must be a valid URL, e.g. http://127.0.0.1:8080/v1"); }

try {
  new LocalAiLLM(embedder, model);
} catch (e) {
  if (/LocalAI Base Path/i.test(e.message)) return promptEnvSetup("LOCAL_AI_BASE_PATH");
  throw e;
}

When it happens

Trigger: Selecting LocalAI as the LLM provider while LOCAL_AI_BASE_PATH is missing from the server env: LocalAI not running or installed yet, wrong port for the installation (common ports vary: 8080 for the binary/docker), or Docker localhost-vs-host networking confusion.

Common situations: Fresh LocalAI deployments where the user configured only the model name; reverse-proxy setups where the public URL was never copied into .env; mixed-up variable names after migrating from LiteLLM/Ollama configs; containerized AnythingLLM pointing at 127.0.0.1 which refers to the container itself.

Related errors

• KoboldCPP must have a valid base path to use for the api.
• No Lemonade API Base Path was set.
• No LMStudio API Base Path was set.
• LiteLLM must have a valid base path to use for the api.
• No embedding base path was set.

AI-assisted analysis of Mintplex-Labs/anything-llm@3aec848f28 (2026-08-18). Data as JSON: /api/errors/c1db983ebc051b39. Report an issue: GitHub.

Appendix: source

Thrown at server/utils/AiProviders/localAi/index.js:13

const { NativeEmbedder } = require("../../EmbeddingEngines/native");
const {
  LLMPerformanceMonitor,
} = require("../../helpers/chat/LLMPerformanceMonitor");
const {
  handleDefaultStreamResponseV2,
  formatChatHistory,
} = require("../../helpers/chat/responses");

class LocalAiLLM {
  constructor(embedder = null, modelPreference = null) {
    if (!process.env.LOCAL_AI_BASE_PATH)
      throw new Error("No LocalAI Base Path was set.");

    this.className = "LocalAiLLM";
    const { OpenAI: OpenAIApi } = require("openai");
    this.openai = new OpenAIApi({
      baseURL: process.env.LOCAL_AI_BASE_PATH,
      apiKey: process.env.LOCAL_AI_API_KEY ?? null,
    });
    this.model = modelPreference || process.env.LOCAL_AI_MODEL_PREF;
    this.limits = {
      history: this.promptWindowLimit() * 0.15,
      system: this.promptWindowLimit() * 0.15,
      user: this.promptWindowLimit() * 0.7,
    };

    this.embedder = embedder ?? new NativeEmbedder();
    this.defaultTemp = 0.7;
  }

View on GitHub (pinned to 3aec848f28)