Mintplex-Labs/anything-llm · error · Error

OpenAI chat: ${this.model} is not valid for chat completion!

Error message

OpenAI chat: ${this.model} is not valid for chat completion!

What it means

Pre-flight model validation in getChatCompletion. isValidChatCompletionModel short-circuits to true for ids containing 'gpt' or starting with 'o'; for any other id it calls openai.models.retrieve(modelName) and returns false if that 404s or errors. This throw fires only for non-gpt/non-o ids that the OpenAI account cannot retrieve.

Source

Thrown at server/utils/AiProviders/openAi/index.js:150

   * @param {string} modelName
   * @param {number} temperature
   * @returns {number}
   */
  #temperature(modelName, temperature) {
    // For models that don't support temperature
    // OpenAI accepts temperature 1
    const NO_TEMP_MODELS = ["o", "gpt-5"];

    if (NO_TEMP_MODELS.some((prefix) => modelName.startsWith(prefix))) {
      return 1;
    }

    return temperature;
  }

  async getChatCompletion(messages = null, { temperature = 0.7 }) {
    if (!(await this.isValidChatCompletionModel(this.model)))
      throw new Error(
        `OpenAI chat: ${this.model} is not valid for chat completion!`
      );

    const result = await LLMPerformanceMonitor.measureAsyncFunction(
      this.openai.responses
        .create({
          model: this.model,
          input: messages,
          store: false,
          temperature: this.#temperature(this.model, temperature),
        })
        .catch((e) => {
          throw new Error(e.message);
        })
    );

    if (!result.output.hasOwnProperty("output_text")) return null;

View on GitHub (pinned to 526360e320)

Solutions

1. List accessible models with the same key: 'openai models list' / GET /v1/models and confirm the id appears.
2. Switch to a gpt- or o- prefixed model, or correct the typo in OPEN_MODEL_PREF.
3. For fine-tunes, ensure the key belongs to the org/project that owns the model.
4. If the model was deprecated, choose the recommended successor.

Example fix

// before - non-gpt id that 404s on models.retrieve
OPEN_MODEL_PREF=my-custom-ft-model

// after - use an accessible id or a gpt- prefix short-circuit
OPEN_MODEL_PREF=gpt-4.1-mini

const isAccessibleOpenAiModel = async (openai, model) => {
  if (model.toLowerCase().includes('gpt') || model.toLowerCase().startsWith('o')) return true;
  try { await openai.models.retrieve(model); return true; } catch { return false; }
};
if (!(await isAccessibleOpenAiModel(openai, modelId)))
  throw new Error(`OpenAI model '${modelId}' is not accessible with this key.`);

const isPresetOpenAiModel = (m) =>
  typeof m === 'string' && (m.toLowerCase().includes('gpt') || m.toLowerCase().startsWith('o'));

When it happens

Trigger: this.model is a non-gpt/non-o id that does not exist, is deprecated, is a fine-tune from another org/project, or the API key lacks access; openai.models.retrieve rejects.

Common situations: Fine-tuned model id scoped to a different organization/project; deprecated model id still saved as OPEN_MODEL_PREF; typo in the model id; new model name not yet retrievable; key without project access.

Related errors

• No base URL was set.
• No token context limit was set.
• No NVIDIA NIM token context limit was set.
• OMLX must have a valid model set.
• No OpenAI API key was set.

AI-assisted analysis of Mintplex-Labs/anything-llm@526360e320 (2026-08-13). Data as JSON: /api/errors/28909f92f1139535. Report an issue: GitHub.