What causes "GenericOpenAI Failed to embed: ${error.message}" in Mintplex-Labs/anything-llm?

Inspect error.message — it is the literal response body from your server

Mintplex-Labs/anything-llm · error · Error

GenericOpenAI Failed to embed

Error message

GenericOpenAI Failed to embed: ${error.message}

What it means

Thrown from GenericOpenAiEmbedder.embedChunks when any single chunk request fails; the loop aborts the whole sequence because partial embeddings would be incomplete. The error object is normalized first (type from error.code/status, message from the response body), so the thrown text after the colon is the upstream endpoint's own message.

Solutions

Inspect error.message — it is the literal response body from your server
For 401s, set GENERIC_OPEN_AI_EMBEDDING_API_KEY (or fix it) to a valid key for that endpoint
For 404s, correct EMBEDDING_BASE_PATH so it is the OpenAI-compatible root (commonly ends in /v1)
For 429s, set GENERIC_OPEN_AI_EMBEDDING_API_DELAY_MS (minimum 500) to throttle between batches
For context-length errors, lower EMBEDDING_MODEL_MAX_CHUNK_LENGTH
Verify EMBEDDING_MODEL_PREF exactly matches a model id the endpoint serves

Example fix

# before
EMBEDDING_BASE_PATH=http://localhost:8080
# 404: SDK posts to /embeddings at server root

# after
EMBEDDING_BASE_PATH=http://localhost:8080/v1
GENERIC_OPEN_AI_EMBEDDING_API_DELAY_MS=500

Defensive patterns

Strategy: try-catch

Validate before calling

// Smoke-test the endpoint before a bulk run: one tiny embedding must round-trip
async function genericOpenAiEmbedderHealthy(openai, model) {
  try {
    const res = await openai.embeddings.create({ model, input: ["ping"] });
    return Array.isArray(res?.data?.[0]?.embedding);
  } catch (e) {
    console.error("Embedding pre-flight failed:", e.status, e.message);
    return false;
  }
}

Try / catch

try {
  const vectors = await embedder.embedTextInput(text);
} catch (e) {
  if (e.message.startsWith("GenericOpenAI Failed to embed:")) {
    const msg = e.message;
    if (/401|unauthorized/i.test(msg)) { /* fix GENERIC_OPEN_AI_EMBEDDING_API_KEY; do not retry */ }
    else if (/429|rate/i.test(msg)) { /* wait, then retry; consider GENERIC_OPEN_AI_EMBEDDING_API_DELAY_MS */ }
    else if (/404|not found/i.test(msg)) { /* fix EMBEDDING_BASE_PATH or EMBEDDING_MODEL_PREF; do not retry */ }
    else throw e;
  } else throw e;
}

Prevention

Set GENERIC_OPEN_AI_EMBEDDING_API_DELAY_MS (>=500) when the backend is rate-limit-prone
Keep EMBEDDING_MODEL_MAX_CHUNK_LENGTH within the target model's context so oversized chunks never reach the API
Remember the loop aborts on first error — after a fix, re-embed the whole document, not the remainder

When it happens

Trigger: 401 from a server that requires a key when GENERIC_OPEN_AI_EMBEDDING_API_KEY is null or wrong; POST {EMBEDDING_BASE_PATH}/embeddings returning 404 because the base path is wrong (missing /v1 or points at a non-OpenAI route); model name in EMBEDDING_MODEL_PREF unknown to the server; 429 rate limiting on small self-hosted or shared endpoints; input text longer than the server's context window.

Common situations: Self-hosted single-threaded backends that 429 under AnythingLLM's batch load; pointing at Ollama's /v1 with a model id that is not pulled; using an OpenAI-compatible facade that does not implement /embeddings; very large documents blowing the max chunk length.

Related errors

AI-assisted analysis of Mintplex-Labs/anything-llm@3aec848f28 (2026-08-18). Data as JSON: /api/errors/360ff021173fe66d. Report an issue: GitHub.

Appendix: source

Thrown at server/utils/EmbeddingEngines/genericOpenAi/index.js:156

          .create({
            model: this.model,
            input: chunk,
          })
          .then((result) => resolve({ data: result?.data, error: null }))
          .catch((e) => {
            e.type =
              e?.response?.data?.error?.code ||
              e?.response?.status ||
              "failed_to_embed";
            e.message = e?.response?.data?.error?.message || e.message;
            resolve({ data: [], error: e });
          });
      });

      // If any errors were returned from OpenAI abort the entire sequence because the embeddings
      // will be incomplete.
      if (error)
        throw new Error(`GenericOpenAI Failed to embed: ${error.message}`);
      allResults.push(...(data || []));
      reportEmbeddingProgress(allResults.length, textChunks.length);
      if (this.apiRequestDelay) await this.runDelay();
    }

    return allResults.length > 0 &&
      allResults.every((embd) => embd.hasOwnProperty("embedding"))
      ? allResults.map((embd) => embd.embedding)
      : null;
  }
}

module.exports = {
  GenericOpenAiEmbedder,
};

View on GitHub (pinned to 3aec848f28)