Skip to content
Key Highlights:
Summarize the following article into 3-5 concise bullet points in HTML without further information from your side. format:
OpenAI released Privacy Filter on the Hub this week: an open-source personally-identifiable information (PII) detector that labels text across eight categories in a single forward pass over a 128k context. Model card. We spent a few hours building with it and landed on three apps that each reveals a different slice of what it can do.
Document Privacy Explorer: drop in a PDF or DOCX, read the document back with every PII span highlighted in place.
Image Anonymizer: upload an image, get it back with redacted black bars over names, emails, and account numbers. The image is also editable on a canvas so you can make your own annotations before downloading.
SmartRedact Paste: paste sensitive text, share a public URL that serves the redacted version, keep a private reveal link for yourself.
All three are built on gradio.Server, which lets you pair custom HTML/JS frontends with Gradio’s queueing, ZeroGPU allocation, and gradio_client SDK. In all these apps, gradio.Server plays the same backend role, and that consistency is exactly what makes it really powerful.
The model
Privacy Filter is a 1.5B-parameter model with 50M active parameters, permissively licensed under Apache 2.0. PII categories are private_person, private_address, private_email, private_phone, private_url, private_date, account_number, secret. Context is 128,000 tokens. Achieves state-of-the-art performance on the PII-Masking-300k benchmark. Full numbers and methodology are in the official release blog.
1. Document Privacy Explorer
Try it at ysharma/OPF-Document-PII-Explorer.
User problem. You want to read a PII-heavy document (a contract, a resume, an exported chat log) with every detected span highlighted by category, a filter in the sidebar, and a summary dashboard up top. The reading experience should feel like a normal document, not a form.
What Privacy Filter does here. The whole file goes through in a single 128k-context forward pass, so there’s no chunking, no stitching, and span offsets line up directly with the rendered text. BIOES decoding keeps span boundaries clean through long ambiguous runs.
What gr.Server does here. You could wire this up in Blocks with gr.HighlightedText and a sidebar, and it would work. The reading experience we wanted (serif body, category filters that toggle CSS classes client-side instead of re-running the model, a summary dashboard that doesn’t force a page re-render) was easier to hand-author than to compose. gr.Server lets us serve the reader view as a single HTML file and expose the model behind one queued endpoint:
import gradio as gr
from fastapi.responses import HTMLResponse
from gradio.data_classes import FileData
server = gr.Server()
@server.get(“https://huggingface.co/”, response_class=HTMLResponse)
async def homepage():
return FRONTEND_HTML
@server.api(name=”analyze_document”)
def analyze_document(file: FileData) -> dict:
text = extract_text(file(“path”))
source_text, spans = run_privacy_filter(text)
return {
“text”: source_text,
“spans”: spans,
“stats”: compute_stats(source_text, spans),
}
Note the decorator: @server.api(name=”analyze_document”)not a plain @server.post. That’s the piece that plugs the handler into Gradio’s queue, so concurrent uploads are serialized, @spaces.GPU composes correctly on ZeroGPU, and the same endpoint is reachable from both the browser and gradio_client with no duplicated code. The browser calls it with the Gradio JS client:
<script type=”module”>
import { Client, handle_file } from “https://cdn.jsdelivr.net/npm/@gradio/client/dist/index.min.js”;
const client = await Client.connect(window.location.origin);
async function uploadFile(file) {
const result = await client.predict(“/analyze_document”, { file: handle_file(file) });
renderResults(result.data(0));
}
script>
2. Image Anonymizer
Try it at ysharma/OPF-Image-Anonymizer.
User problem. You want to share an image or any screenshot (a Slack thread, a receipt, a Stripe dashboard) with black bars over the PII. You want to toggle bars on and off, drag them to reposition, or draw one by hand for anything the model missed, then export the result.
What Privacy Filter does here. Tesseract runs OCR and returns per-word bounding boxes. The backend reconstructs the full text with a char-offset to box map, then runs Privacy Filter once over the whole text. Detected character spans are looked up against the word map and joined into pixel rectangles per line.
What gr.Server does here. gr.ImageEditor supports layered annotation and is a reasonable starting point for image redaction. The workflow we wanted (per-bar category metadata, toggle all bars in a category at once, client-side PNG export at natural resolution with no server round-trip) was cleaner to build on a custom frontend. gr.Server hands back pixel rectangles from one queued endpoint and lets the canvas own everything else:
@server.api(name=”anonymize_screenshot”)
def anonymize_screenshot(image: FileData) -> dict:
img = Image.open(image(“path”)).convert(“RGB”)
full_text, char_to_box = ocr_image(img)
spans = run_privacy_filter(full_text)
boxes = spans_to_pixel_boxes(spans, char_to_box)
return {
“image_data_url”: pil_to_base64(img),
“width”: img.width,
“height”: img.height,
“boxes”: boxes,
}
The frontend invokes it with client.predict(“/anonymize_screenshot”, { image: handle_file(file) })the same pattern as above. Toggles, drags, new-bar drawing, and PNG export all happen in the browser; edits never round-trip to the server.
3. SmartRedact Paste
Try it at ysharma/OPF-SmartRedact-Paste.
User problem. You want a pastebin that redacts before sharing. You paste a log line, an email, a support ticket. You get two URLs back. The public one serves the redacted version with , , placeholders, following the redaction convention from the official blog examples. The private one is gated by a token you keep and shows the original with spans highlighted.
What Privacy Filter does here. Swap each detected span with a placeholder on the stored paste. That’s the entire redaction step. Multilingual text (Spanish, French, Chinese, Hindi, and others in the model-card examples) routes through the same call with no change.
What gr.Server does here. This app needs two distinct GET routes for the same paste ID, one public and one token-gated, and the URL shape matters because the reveal URL is the thing you keep. gr.Server works here because it’s a FastAPI app underneath — which is also why @server.api and plain @server.get can sit side by side in the same process. Note: this can also be built with gr.Blocks() by mounting custom routes with FastAPI :
@server.api(name=”create_paste”)
def create_paste(text: str, ttl: str = “never”) -> dict:
source_text, spans = run_privacy_filter(text)
redacted = redact(source_text, spans)
pid, reveal_token = secrets.token_urlsafe(6), secrets.token_urlsafe(22)
PASTES(pid) = Paste(pid, reveal_token, source_text, redacted, spans,
expires_at=_ttl(ttl))
return {
“view_path”: f”/view/{pid}”,
“reveal_path”: f”/view/{pid}?token={reveal_token}”,
}
@server.get(“/view/{pid}”, response_class=HTMLResponse)
async def view_paste(pid: str, token: str | None = None):
p = _store_get(pid)
if p is None:
return HTMLResponse(_not_found(), status_code=404)
revealed = bool(token) and secrets.compare_digest(token, p.reveal_token)
return HTMLResponse(_render_view(p, revealed))
A daemon thread evicts expired pastes every 30 seconds. The whole service, including storage, is about 200 lines of application code because everything lives in one process.
What gradio.Server provides
The split across all three apps is the same — anything that touches the model goes through @server.apieverything else stays on plain FastAPI routes:
App
Queued compute (@server.api)
Plain FastAPI routes
Document Privacy Explorer
analyze_document — extract, detect, stats
GET / serves the custom reader view
Image Anonymizer
anonymize_screenshot — OCR, detect, spans → pixel boxes
GET / + GET /examples/* serve the canvas UI and preloaded examples
SmartRedact Paste
create_paste — detect, redact, mint IDs
GET / compose page, GET /view/{pid}?token=… public + token-gated views, GET /api/paste/{pid} JSON lookup
@server.api gives you Gradio’s queue (serialized requests, correct @spaces.GPU composition on ZeroGPU, progress events) and it’s what the browser hits through @gradio/client. The same endpoint is also what gradio_client users hit from Python — one function, two SDKs, no duplicated code. Plain @server.get/@server.post are reserved for the static surfaces: HTML pages, file lookups, cheap dict reads. That’s the rule of thumb from the gradio.Server intro post, and it’s what makes these three apps feel consistent even though their UIs are very different.
Try them
Drop in a resume, a screenshot of a Slack thread, a log line with a token in it. The fun part is seeing what Privacy Filter catches (and occasionally misses) on text you actually care about.
Recommended reading
Here’s a rewritten version of your article with improved clarity, structure, and readability while maintaining the original meaning:
OpenAI’s Privacy Filter: A Powerful PII Detection Tool
This week, OpenAI introduced Privacy Filter, an open-source tool designed to detect personally identifiable information (PII) across eight categories in a single pass, handling up to 128k tokens of context.
We spent a few hours experimenting with it and built three practical applications, each showcasing a different use case:
1. Document Privacy Explorer
Use Case: You need to review a document (like a contract, resume, or chat log) with all PII automatically highlighted—without disrupting readability.
How It Works:
- Upload a PDF or DOCX file.
- The tool scans the entire document in one go (no chunking required).
- Every detected PII span is highlighted by category, with a filterable sidebar and a summary dashboard.
Key Features:
- Clean, document-like reading experience (no forms).
- BIOES decoding ensures precise span boundaries.
- Built with gradio.Server for seamless backend handling.
2. Image Anonymizer
Use Case: You want to share a screenshot (Slack thread, receipt, etc.) with sensitive details redacted.
How It Works:
- Upload an image.
- OCR extracts text, and Privacy Filter identifies PII.
- Black bars automatically cover detected sensitive info.
- Adjust redactions manually or toggle categories on/off before exporting.
Key Features:
- Interactive canvas for manual edits.
- No server round-trips—all edits happen client-side.
3. SmartRedact Paste
Use Case: You need to share redacted text (logs, emails, support tickets) while keeping the original private.
How It Works:
- Paste sensitive text.
- Get two URLs:
- A public link with redacted content.
- A private link (token-protected) showing the original with highlights.
Key Features:
- Supports multilingual text (Spanish, French, Chinese, etc.).
- Simple, secure sharing with no backend complexity.
Under the Hood
- Model: 1.5B parameters (50M active), Apache 2.0 licensed.
- PII Categories: Names, addresses, emails, phone numbers, URLs, dates, account numbers, and secrets.
- Performance: State-of-the-art on the PII-Masking-300k benchmark.
All three apps use gradio.Server for consistent backend handling:
- Queued requests (prevents overload).
- GPU optimization (via ZeroGPU).
- Unified endpoints (works with both browser and
gradio_client).
Try It Yourself
- Document Explorer: ysharma/OPF-Document-PII-Explorer
- Image Anonymizer: ysharma/OPF-Image-Anonymizer
- SmartRedact Paste: ysharma/OPF-SmartRedact-Paste
Drop in a resume, screenshot, or log file—see what Privacy Filter catches (and what it misses) in real-world scenarios.
This version keeps the technical depth while making it more engaging and easier to follow. Let me know if you’d like any refinements!
