n8n-io/n8n · warning · UserError

Additional Model Request Fields must be valid JSON

Error message

Additional Model Request Fields must be valid JSON

What it means

Identical in spirit to error 682 but in the Bedrock Chat node: options.additionalModelRequestFields is a free-text JSON string; if it is non-empty, non-'{}', and fails jsonParse, a UserError (warning level) is thrown. Bedrock's Converse API accepts arbitrary model request overrides as a JSON object.

Source

Thrown at packages/@n8n/nodes-langchain/nodes/llms/LmChatAwsBedrock/LmChatAwsBedrock.node.ts:432

		const guardrail = options.guardrail?.values;
		if (guardrail?.guardrailIdentifier) {
			modelConfig.guardrailConfig = {
				guardrailIdentifier: guardrail.guardrailIdentifier,
				// AWS requires a version whenever guardrailConfig is sent. The field may be blank,
				// whitespace (e.g. from an expression), or absent (collection defaults only
				// materialize on add); fall back to the working draft.
				guardrailVersion: guardrail.guardrailVersion?.trim() || 'DRAFT',
				...(guardrail.trace ? { trace: guardrail.trace } : {}),
			};
		}

		const additionalFields = options.additionalModelRequestFields?.trim();
		if (additionalFields && additionalFields !== '{}') {
			let parsed: DocumentType;
			try {
				parsed = jsonParse<DocumentType>(additionalFields);
			} catch {
				throw new UserError('Additional Model Request Fields must be valid JSON', {
					level: 'warning',
				});
			}
			modelConfig.additionalModelRequestFields = parsed;
		}

		const model = new ChatBedrockConverse(modelConfig);

		return {
			response: model,
		};
	}
}

View on GitHub (pinned to 5ac6606e81)

Solutions

1. Open 'Additional Model Request Fields' and paste valid JSON, e.g. {"inferenceConfig": {"maxTokens": 512}}.
2. Run the string through a JSON linter.
3. If templated, ensure the expression returns a valid JSON string or use {} / leave empty.
4. Consult the Bedrock Converse API docs for the exact key names allowed.

Example fix

// before
options.additionalModelRequestFields = "inferenceConfig: {maxTokens: 512}"
// after
options.additionalModelRequestFields = "{\"inferenceConfig\": {\"maxTokens\": 512}}"

const raw = options.additionalModelRequestFields?.trim();
if (raw && raw !== '{}') {
  try { JSON.parse(raw); }
  catch (e) { throw new Error(`Fix JSON: ${(e as Error).message}`); }
}

const isJsonString = (s: string): boolean => { try { JSON.parse(s); return true; } catch { return false; } };

// Pre-validate at save; runtime catch only rewraps.

When it happens

Trigger: User enters malformed JSON in 'Additional Model Request Fields' on the Bedrock Chat node — unquoted keys, trailing commas, smart quotes, or a templated expression resolving to invalid JSON.

Common situations: Pasting from Bedrock docs with single quotes; expression returning undefined; hand-editing; copy-paste with mangled characters.

Related errors

• Additional Model Request Fields must be valid JSON
• The value in the "Extra Body" field is not valid JSON
• Additional Model Request Fields must be a JSON object
• No model selected. Please choose a model.
• The value in the "Extra Body" field must be a JSON object

AI-assisted analysis of n8n-io/n8n@5ac6606e81 (2026-08-12). Data as JSON: /api/errors/9a46001c1f3c2291. Report an issue: GitHub.