byteit 1.0.0

AI-powered document intelligence platform - Turn your data into structured data with a single line of code.

ByteIT Python SDK

Python client for ByteIT — AI-powered document parsing. Extract structured text from PDFs, Word files, images, and more with a single API call.

---

Installation

pip install byteit

Requires Python 3.8+ and an API key from byteit.ai.

---

Quick Start

from byteit import ByteITClient

client = ByteITClient(api_key="your_api_key")
result = client.parse("document.pdf")
print(result.decode())

Returns raw bytes. Pass output="result.md" to save directly to disk.

---

Usage

Parse and save

# Returns bytes
result = client.parse("invoice.pdf", result_format="json")

# Save to file
client.parse("invoice.pdf", result_format="md", output="invoice.md")

Output formats: md (default), txt, json, html

Async (non-blocking)

Submit a job and check back later — useful for large files or batch workflows.

# Submit without waiting
job = client.parse_async("document.pdf")

# Poll status
status = client.get_job_status(job.id)
# status.processing_status: "pending" | "processing" | "completed" | "failed"

# Download when ready
if status.is_completed:
    result = client.get_job_result(job.id)

Job management

for job in client.get_jobs():
    print(f"{job.id}  {job.processing_status}  {job.result_format}")

Processing options

from byteit import ProcessingOptions

result = client.parse(
    "document.pdf",
    processing_options=ProcessingOptions(languages=["de", "en"], page_range="1-5"),
)

Or pass a plain dict:

result = client.parse("doc.pdf", processing_options={"languages": ["de"]})

API key from environment

import os
client = ByteITClient(api_key=os.environ["BYTEIT_API_KEY"])

Context manager

with ByteITClient(api_key="your_key") as client:
    result = client.parse("doc.pdf")

---

Supported File Types

Documents	Images
PDF .pdf	PNG .png
Word .docx	JPEG .jpg .jpeg
PowerPoint .pptx	TIFF .tiff
HTML .html	BMP .bmp
Markdown .md
Plain text .txt
JSON .json
XML .xml

---

Error Handling

All exceptions inherit from ByteITError.

from byteit.exceptions import (
    AuthenticationError,
    ValidationError,
    RateLimitError,
    JobProcessingError,
    ByteITError,
)

try:
    result = client.parse("document.pdf")
except AuthenticationError:
    print("Invalid API key")
except ValidationError as e:
    print("Bad request:", e.message)
except RateLimitError:
    print("Rate limit hit — retry later")
except JobProcessingError as e:
    print("Processing failed:", e.message)
except ByteITError as e:
    print("Unexpected error:", e.message)

Exception	When raised
AuthenticationError	Invalid or missing API key
APIKeyError	API key rejected (403)
ValidationError	Bad request parameters
ResourceNotFoundError	Job not found
RateLimitError	Rate limit exceeded
JobProcessingError	Job failed during processing
ServerError	Server-side error (5xx)

---

API Reference

ByteITClient(api_key)

Method	Description
parse(input, ...)	Parse a document, block until complete, return bytes
parse_async(input, ...)	Submit a job, return Job immediately
get_job_status(job_id)	Get current Job status
get_job_result(job_id)	Download result as bytes
get_jobs()	List all jobs as list[Job]

parse(input, output=None, processing_options=None, result_format="md") → bytes

Param	Type	Description
input	str | Path | InputConnector	File to parse
output	str | Path | None	Save result to disk (optional)
processing_options	ProcessingOptions | dict | None	Languages, page range, etc.
result_format	str	"md", "txt", "json", "html"

parse_async(input, processing_options=None, result_format="md") → Job

Same parameters as parse, minus output. Returns a Job without waiting.

Job properties

Property	Type	Description
id	str	Unique job identifier
processing_status	str	pending / processing / completed / failed
result_format	str	Output format
is_completed	bool	True when result is ready
is_failed	bool	True if job failed
metadata	DocumentMetadata	Filename, page count, language, etc.

---

Notebook Integration

Results are automatically rendered when running in Jupyter:

• md → rendered Markdown
• html → rendered HTML
• json → interactive tree
• txt → code block

To disable auto-display, pass output="file.md".

---

Resources

• Studio: studio.byteit.ai — Process and test with a graphical user interface.
• Colab notebook: Quick demo
• Pricing: byteit.ai/pricing — 1,000 free credits
• Support: byteit.ai/support
• LinkedIn: ByteIT on LinkedIn

---

Licensed under Apache 2.0. © 2026 ByteIT GmbH.