Mintplex-Labs/anything-llm · error · Error

Ollama::getChatCompletion failed to communicate with Ollama.

Error message

Ollama::getChatCompletion failed to communicate with Ollama. ${this.#errorHandler(e).message}

What it means

Catch-all wrapper around any rejection from this.client.chat(...) in the non-streaming path. It routes the error through #errorHandler, which either rewrites 'fetch failed' into the friendly unreachable message (error 263) or returns the original error; the result is prefixed with 'Ollama::getChatCompletion failed to communicate with Ollama.' This is the umbrella error for all non-network Ollama chat failures.

Source

Thrown at server/utils/AiProviders/ollama/index.js:298

            num_ctx: this.promptWindowLimit(),
          },
        })
        .then((res) => {
          let content = res.message.content;
          if (res.message.thinking)
            content = `<think>${res.message.thinking}</think>${content}`;
          return {
            content,
            usage: {
              prompt_tokens: res.prompt_eval_count,
              completion_tokens: res.eval_count,
              total_tokens: res.prompt_eval_count + res.eval_count,
              duration: res.eval_duration / 1e9,
            },
          };
        })
        .catch((e) => {
          throw new Error(
            `Ollama::getChatCompletion failed to communicate with Ollama. ${this.#errorHandler(e).message}`
          );
        })
    );

    if (!result.output.content || !result.output.content.length)
      throw new Error(`Ollama::getChatCompletion text response was empty.`);

    return {
      textResponse: result.output.content,
      metrics: {
        prompt_tokens: result.output.usage.prompt_tokens,
        completion_tokens: result.output.usage.completion_tokens,
        total_tokens: result.output.usage.total_tokens,
        outputTps:
          result.output.usage.completion_tokens / result.output.usage.duration,
        duration: result.output.usage.duration,
        model: this.model,

View on GitHub (pinned to 526360e320)

Solutions

1. Read the suffix after 'failed to communicate with Ollama.' to get the true cause.
2. If the suffix is the unreachable message, follow error 263's connectivity steps.
3. Run 'ollama list' on the server and confirm this.model is present; 'ollama pull' if missing.
4. For context/OOM errors, lower OLLAMA_MODEL_TOKEN_LIMIT or reduce injected context.

Example fix

// before
.catch((e) => {
  throw new Error(`Ollama::getChatCompletion failed to communicate with Ollama. ${this.#errorHandler(e).message}`);
})

// after - distinguish network vs model errors for callers
.catch((e) => {
  const friendly = this.#errorHandler(e);
  const wrapped = new Error(`Ollama::getChatCompletion failed: ${friendly.message}`);
  wrapped.code = e?.cause?.code || (e.message === 'fetch failed' ? 'ECONNREFUSED' : 'OLLAMA_ERROR');
  throw wrapped;
})

const ensureOllamaModel = async (client, model) => {
  const { models = [] } = await client.list().catch(() => ({ models: [] }));
  if (!models.some((m) => m.name === model))
    throw new Error(`Ollama model '${model}' is not pulled. Run: ollama pull ${model}`);
};
await ensureOllamaModel(llm.client, llm.model);

const isOllamaCommError = (e) =>
  !!e && /failed to communicate with Ollama/.test(e.message);

try {
  return await llm.getChatCompletion(messages, { temperature });
} catch (e) {
  if (e.message.includes('could not be reached')) { /* connectivity path */ }
  if (e.message.toLowerCase().includes('model not found')) { /* pull/fix model */ }
  throw e;
}

When it happens

Trigger: Model not pulled on the Ollama server; model name typo; num_ctx/prompt exceeding the model's capacity; GPU OOM during inference; Ollama internal 500; connection refused (suffix becomes the fetch-failed message); invalid options payload.

Common situations: User selected a model in the UI that was never 'ollama pull'ed; quantized model OOMs on the available VRAM; chat history grew past the context window; Ollama restarted mid-session; wrong model id casing.

Related errors

• Your Ollama instance could not be reached or is not respondi
• Ollama::getChatCompletion text response was empty.
• Stream returned undefined chunk. Aborting reply - check mode
• ${e.message}
• ${e.message}

AI-assisted analysis of Mintplex-Labs/anything-llm@526360e320 (2026-08-13). Data as JSON: /api/errors/b240b4e2cc51fe25. Report an issue: GitHub.