n8n-io/n8n · warning · Error

Unsupported binary file — only PNG, JPEG, GIF, WebP and PDF

Error message

Unsupported binary file — only PNG, JPEG, GIF, WebP and PDF are readable

What it means

Thrown by read_file when the file content is detected as binary (via isLikelyBinaryContent — presence of null bytes or non-decodable UTF-8) AND the file extension is not one of the supported binary types (PNG, JPEG, GIF, WebP, PDF — detected by detectSupportedBinaryFile). Supported image and PDF files are returned as binary results; unsupported binary formats are rejected.

Source

Thrown at packages/@n8n/computer-use/src/tools/filesystem/read-file.ts:56

		const resolvedPath = await resolveReadablePath(dir, filePath);

		const stat = await fs.stat(resolvedPath);
		if (stat.size > MAX_FILE_SIZE) {
			throw new Error(
				`File too large: ${stat.size} bytes (max ${MAX_FILE_SIZE} bytes). Use searchFiles for specific content.`,
			);
		}

		const fileContent = await fs.readFile(resolvedPath);
		const buffer = Buffer.isBuffer(fileContent) ? fileContent : Buffer.from(fileContent);

		const binaryType = detectSupportedBinaryFile(filePath);
		if (binaryType) {
			return buildBinaryResult(resolvedPath, buffer, binaryType);
		}

		if (isLikelyBinaryContent(buffer)) {
			throw new Error('Unsupported binary file — only PNG, JPEG, GIF, WebP and PDF are readable');
		}

		return buildTextResult(filePath, buffer.toString('utf-8'), startLine, maxLines);
	},
};

function buildTextResult(
	filePath: string,
	content: string,
	startLine: number | undefined,
	maxLines: number | undefined,
): CallToolResult {
	const allLines = content.split('\n');
	const lines = maxLines ?? DEFAULT_MAX_LINES;
	const start = startLine ?? 1;
	const startIndex = Math.max(0, start - 1);
	const slicedLines = allLines.slice(startIndex, startIndex + lines);
	const truncated = allLines.length > startIndex + lines;

View on GitHub (pinned to 5ac6606e81)

Solutions

1. Convert the file to a supported image format (PNG/JPEG/GIF/WebP) or PDF before reading
2. Use shell tools to extract text content (e.g. unzip + read the inner XML for .docx)
3. Use a dedicated parser or library for the specific binary format

import { isLikelyBinaryContent } from './fs-utils';

const SUPPORTED_BINARY_EXTENSIONS = new Set(['.png', '.jpg', '.jpeg', '.gif', '.webp', '.pdf']);

async function isReadableTextOrSupportedBinary(filePath: string): Promise<boolean> {
  const ext = filePath.toLowerCase().match(/\.[^.]+$/)?.[0] ?? '';
  if (SUPPORTED_BINARY_EXTENSIONS.has(ext)) return true;
  const buffer = await fs.readFile(filePath);
  return !isLikelyBinaryContent(buffer);
}

function isUnsupportedBinaryError(e: unknown): boolean {
  return e instanceof Error && e.message.startsWith('Unsupported binary file');
}

When it happens

Trigger: Reading a .exe, .zip, .tar, .gz, .docx, .xlsx, .so, .class, .woff, or any other binary file that is not PNG/JPEG/GIF/WebP/PDF. The null-byte or non-UTF-8 content triggers isLikelyBinaryContent, and detectSupportedBinaryFile returns null for the extension.

Common situations: Agent tries to read a compiled binary, a compressed archive, an Office document, or a font file.

Related errors

• File too large: ${stat.size} bytes (max ${MAX_FILE_SIZE} byt
• oldString not found in file: ${filePath}
• Access denied: "${excludedSegment}" is excluded from filesys
• File too large: ${stat.size} bytes (max ${MAX_FILE_SIZE} byt
• Pattern "${pattern}" escapes the base directory

AI-assisted analysis of n8n-io/n8n@5ac6606e81 (2026-08-12). Data as JSON: /api/errors/d85b6b07500cbe28. Report an issue: GitHub.