Premium
WhatsApp Filter

Understand the workspace before your first run.

Premium WhatsApp Filter (PWF) is a Windows desktop application for checking phone numbers against WhatsApp, collecting available profile information, downloading avatars, and analyzing images. It includes local checking, a separate Quick Checker, channel testing, file preparation tools, and a connection to the PWF cloud platform.

What this guide covers

This manual documents the user-facing workflows and their implementation in the supplied source snapshot, reviewed on September 10, 2026. The assembly version is 23.5.0.0; some interface labels display different version text. Use the version reported by your installed build when contacting support.

The instructions below are based on source inspection. Live authentication, WhatsApp connectivity, cloud orders, and paid services were not exercised while preparing this document.

Prepare the desktop application and authenticate your license.

Before opening PWF

• Use the complete Windows application package. Keep its supporting libraries, web assets, database resources, and AI worker files together.
• The project targets .NET Framework 4.8.1 and uses Microsoft Edge WebView2 for its interface. The required runtime and matching native dependencies must be available.
• Provide a writable application/data location and enough free disk space for results, downloaded avatars, and exported files.
• Internet connectivity is needed for license verification, WhatsApp checking, avatar downloads, and cloud features. Local image analysis uses local analysis engines once the application is available.

No verified minimum RAM, CPU, or throughput specification is defined by this guide. Resource use depends on channel concurrency, worker counts, images, and list size. The current startup code may request Windows elevation.

Activate a license

1. Launch the supplied executable and wait for the License window to load.
2. Enter your key in License Key. Enable Remember license if you want the application to retain it on this device.
3. Select Authenticate, or press Enter in the key field.
4. Review the returned status, key type, activation date, and expiry.
5. When activation succeeds, select Continue.

The license is associated with hardware information. If authentication fails, check the exact error, your connection, the key, and the device binding. A copied application folder does not by itself transfer activation to another computer.

Find the right screen for each task.

The main table supports paging, page-size selection, filtering, and sorting. Status indicators distinguish loaded input, connection state, elapsed time, and processing activity. Changing the visible page does not limit the underlying dataset.

A complete workflow from input to exported results.

1. Prepare a small number file. Use a TXT list with one international number per line, including country code. Keep a copy of your original file.
2. Import your channels. Open Channels → Import. Choose the correct Personal or Business type for the imported accounts.
3. Check channel health. Use Test All or the separate Channel Checker. Read failures before starting a number run.
4. Load numbers. Open Checker → Select Numbers, or drop a TXT/CSV file onto the number input area.
5. Choose processing options. Review the filtering profile, run mode, thread count, avatar download option, and any required AI analysis.
6. Start the run. Select Start in the main Checker. Watch the number counters and the download/AI queues.
7. Review sample results. Inspect profiles, image availability, and unknown values. If connections fail or results look implausible, stop and inspect Logs.
8. Export. Open Export Data, choose filters and format, verify Matching rows, and save the output.

Stop an active run before clearing results, importing replacement datasets where blocked, testing channels where blocked, or using data maintenance commands. Closing during a run may discard work that has not yet been saved.

Use consistent input so parsing does not distort your results.

The main Checker accepts TXT and CSV files. The most portable format is one number per line with country code. For CSV/TSV-style input, put the number in the first column. This is a simple line-based importer, not a general spreadsheet importer.

+1 202 555 0100
44 7700 900123
12025550101,Example contact

Illustrative formatting examples only; no checking result is implied.

Do not assume every tool and the cloud service use the same parser. For moving files between modules, use clean digits-only international numbers, one per line. Number-format validity and WhatsApp registration are separate checks.

Manage the accounts used to perform local checks.

A channel is an imported account/session used by the checker. It is separate from the target phone numbers you want to check. A channel file contains account material, not simply a contact list.

Import channels

1. Open Channels and select Import.
2. Choose the account type that matches the batch: Personal or Business.
3. Select the supplied TXT/CSV channel file.
4. Review the import counts and test results.

The main importer reads this field order:

number,hash,privateKey,type

The number and Base64-formatted hash are required by the parser; the private-key field can be absent at parse time. Parser acceptance does not guarantee a usable account. In the current main importer, the type chosen in the import dialog applies to the whole batch, even when a row includes a different fourth field. Existing channel numbers are treated as duplicates.

Channel actions

Working, Connecting, Cooldown, Banned, Dead, and other status labels describe different stages or outcomes. A cooldown means the scheduler is resting a channel; it is not the same as a failed login.

Test a batch before adding working accounts to the main workspace.

1. Open Channels → Channel Checker and import a channel file.
2. Confirm or set the correct Business/Personal type.
3. Review Timeout s, the Quick test setting, and Threads. These control test timing and concurrency; they are not the main number-filtering settings.
4. Select Test All or Selected. Use Stop to interrupt the test.
5. Review the Status, Attempts, Time ms, and Details columns.
6. Use Working under Export to save working channels, or To Main to pass working results to the main application.

Use the status chips to inspect each group. An inconclusive result should not automatically be treated as a banned account or a successful login.

Understand what each processing control changes.

Channels contains the main filtering configuration. The presets are labeled Safe, Balanced, Fast, and Custom. These are application presets; none guarantees that accounts will avoid restrictions.

If progress pauses with Sync Avatar Download enabled, inspect DL and AI pending counts and the worker status. The checker can be waiting for analysis rather than for WhatsApp responses. Avoid changing many settings at once when diagnosing a problem.

Read the dataset and stop at a defined matching count.

Main counters

Total and WhatsApp summarize checking outcomes; Non-WA and Business provide additional counts. Avatars describes image availability/download progress. Gender and Objects summarize completed analysis. Pending separates unfinished download and AI work. Filtered tracks processed input. These are different stages, so their counts need not match.

Inspect records

Use chips such as All, Avatar, Male, Female, Business, 50+ M, 50+ F, Flowers, Dogs, and Cats to narrow the table. Available columns include Number, Avatar, Gender, Age, Faces, Objects, Country, Carrier, Type, and Status. Open Profile Details for additional carrier, line-type, formatting, and validity information when available.

The row menu offers Copy Number, Copy Image URL, Open Image, View Profile, Mark Selected, and Delete Row. Deleting a result changes the stored dataset; switching a filter chip only changes the view. Country/carrier information depends on the available lookup data and service responses.

Create a Smart Target

1. Select Add Target and enter a descriptive name.
2. Set the relevant gender, age range, account type, avatar, and object conditions.
3. Set Stop when matches reach to your required matching count.
4. Choose the response: Auto-stop the filter, sound, toast, and/or confetti.
5. Save the target. Use its controls to pause/resume, edit, or remove it.

A target matches the records and analysis currently available to the application. Late image results and in-flight batches can affect when a target is reached. Treat the threshold as a stop trigger, not an exact final export-size guarantee.

Analyze a folder independently of a number-checking run.

1. Open Image Analysis and select Pick Image Folder.
2. Enable Include subfolders if the input should cover nested directories.
3. Select Detect Gender, Detect Objects, or both.
4. Use Analyze pending when you want to focus on unfinished analysis, then select Start.
5. Watch the Images, Gender, Objects, and Pending counters.
6. Filter the table and open Image Details to inspect the file, analysis values, status, and path.
7. Use the image-specific Export page to save the relevant results.

The face analyzer produces estimated age/gender information and a face count. The object analyzer supplies detected content labels. Images without a suitable detected face can remain unknown. Image quality, multiple people, partial faces, and non-photographic avatars can all affect results.

Image result exports

The image export screen offers general filters, Export by Object, and Export by Gender / Age / Faces. Formats include TXT, CSV, an image folder, and XLSX with images. Object filters include any detected object, no object detected, selected labels, and an exclude-object switch where shown. Leave Faces count blank to avoid restricting face count.

For a local folder, source filenames/identifiers belong to image records; do not assume they are verified phone numbers. Review a sample export before passing it to a number-based system.

Manage the local engines that process images.

The Image Analysis Workers page and App Settings expose the face and object pools. The pools are shared with the main number Checker.

More workers do not guarantee faster completion. Disk access, image download speed, available memory, and CPU can become the bottleneck. Worker cards can show progress and recycling/restart information. A recycled worker is not automatically an application failure.

The source supports a native face engine and a fallback external face process, plus an external object-analysis worker. Missing or incompatible engines should be diagnosed through Logs and the worker test. Keep the matching worker executables, models, and libraries from the application package together.

Build a filtered output with the right fields and format.

1. Open Export Data.
2. Choose a Type: All, WhatsApp, Non-WhatsApp, With Photo, No Photo, Business, Not Filtered, or Unknown Gender.
3. Optionally restrict gender, age range, face count, country, and carrier.
4. Choose whether phone numbers should include a leading +.
5. Choose an output format, splitting options, and file prefix.
6. Use Refresh Lists when country/carrier choices are stale. Review Matching rows and Files to create.
7. Select Export and choose the destination. Wait for the completion result before opening the output.

Split files

Rows per file offers a single file or fixed-size parts, from 100 through 10,000 rows in the current interface. Set a meaningful file prefix and review the predicted file count. Splitting controls output file size; it does not change which records match.

Export by Object

Choose a detected object from the list, or type a custom object name such as potted plant. A typed name overrides the list. Available outputs are phone numbers in TXT, avatar images, or XLSX with avatars. Refresh the object list after new analysis and review the match count.

Use the separate list or sequential-generator workflow.

Open Quick Checker from the sidebar or with Ctrl+Q. Its channels, generator/list state, stored results, and exports form a separate workspace. It uses its own SQLite store; do not assume clearing or exporting the main Checker also clears or exports Quick Checker.

Run a check

1. Import channels in the Channels tab and select the correct account type.
2. Load proxies and enable Proxy if required by your configuration.
3. Either use Load list… to supply numbers, or enter a Start Number for sequential generation.
4. Set Target WA Count and review the generator/list status.
5. Select a preset: Conservative, Balanced, or Aggressive. Review Batch, Threads, Login Tries, Delay, and cooldown settings.
6. Choose Skip filtered, Fetch avatars, and Download avatars as needed.
7. Select Start. Use Pause, Resume, and the run's stop controls as appropriate.
8. Review Results, then open Export.

Fetch avatars and Download avatars are distinct: requesting an avatar URL does not necessarily save its image locally. The generator displays Generated, WA Found, and Current #. A generated sequence is candidate input, not a verified contact list.

Inspect results and channels

Results include Number, State, Type, and Avatar. Use WA only, Avatars first, search, and pagination. The channel table includes proxy, type, status, processed count, WA count, speed, and last-seen information. Failed Channels summarizes failed or banned channels from a run.

The channel context menu can Clear ban, Mark banned (24h), or Clear ALL bans. Clearing a local ban flag only changes local tracking; it does not remove a restriction imposed by WhatsApp.

Export

Data Export provides WA only, With avatar, Personal only, and Business only filters. The visible choices include TXT, CSV, JSON, and Numbers only. Review the displayed export count before saving. Use the Quick Checker exporter for these results rather than the main Export Data page.

Create candidate lists from a start number, prefixes, or country data.

Open Tools with Ctrl+T. Generation prepares files; it does not verify number allocation, ownership, reachability, or WhatsApp registration.

Country Gen also exposes AI-ranked network age information. Treat these rankings as generated guidance rather than verified subscriber demographics or allocation dates. Prefix datasets and carrier lookup results can be incomplete or become outdated.

Large quantities can create large files and long-running jobs. Check the preview before generating. Use the operation's Stop control when available and inspect whether the output is complete before reusing it.

Prepare existing lists for checking or handoff.

File Merger

1. Add Files or paste input. Remove unwanted entries and review the list.
2. Use Preview or Analyze Overlap when you need to inspect the inputs.
3. Choose Remove Duplicates, Remove Empty Lines, Trim Lines, and/or Shuffle Output.
4. Review Total Lines and select Merge & Save.

Analyze

Import a list, choose Auto or a fixed Prefix length (2–8), and select Analyze. The results summarize Total Numbers, Matched, Unmatched, Countries, Carriers, and Duplicates. Prefix Breakdown shows country, code, carrier, prefix, count, percentage, and maximum number. Export CSV, TSV, JSON, or Excel; use selected-row or maximum-number exports where appropriate.

Matched in this tool refers to matching prefix data, not WhatsApp registration. Prefix-based carrier classification should not be read as a guaranteed current network assignment.

File Splitter

Paste text or upload a file, choose Lines per file and a File name prefix, then choose individual output files or ZIP. Review Total Lines, Files to create, and Lines per file. Select Split Now, then use Open Folder to inspect completed files.

Remove Duplicates

Paste or upload input. Configure Trim whitespace, Remove empty lines, Ignore case, and Normalize phone numbers. Normalization strips formatting such as +, spaces, and dashes for comparison. Select Remove Duplicates & Save and review Total, Unique, and Removed. The selected normalization settings determine which differently formatted values count as duplicates.

Plus Formatter

Paste or upload numbers. Select Add + Sign or Remove + Sign, and optionally Remove leading zeros (00). Select Format Numbers and inspect Total, Changed, and Unchanged. This tool changes formatting; it does not add a missing country code or verify a number.

Connect an account and manage remote check orders.

The WS Platform section is a remote service integration, separate from the local channel engine. A cloud check submits numbers to the platform and may consume account balance. Local channel imports do not replace a platform access token.

Connect

1. Obtain the Access Token from your platform account's My Profile → Security Settings, as directed by the app.
2. Paste it into Connect to PWF Platform.
3. Optionally enable Remember on this device; the integration offers encrypted local token storage.
4. Select Connect and review Current Balance, Available Services, and Recent Orders.

Create an order

1. Open New Check.
2. Choose a numbers file. The screen documents TXT, one number per line, with a maximum of 1,000,000 numbers. The file picker also accepts CSV; clean one-number-per-line TXT is the least ambiguous input.
3. Select an available service and review the number count and estimated cost.
4. Select Create Check Order when you intend to submit the data and incur the service cost.
5. Use My Orders to monitor status.

The interface removes blank lines and duplicates before its pricing estimate. The server determines final cost and service availability. This manual does not fix prices, balance requirements, or turnaround times.

Download results

Refresh My Orders manually or enable Auto-refresh (30s). Use its page-size and paging controls for larger histories. When downloads are available, choose All Results ZIP, WhatsApp Activated TXT, Unregistered TXT, Results CSV, or Original File as uploaded. Verify the order and result type before saving.

Disconnect ends the application's current platform connection. Protect access tokens as account credentials and avoid including them in screenshots or shared documentation.

Customize the interface and save the changes.

Open App Settings. Use its category tabs to jump to a group, then select Save All or Ctrl+S. Some changes apply immediately, worker counts require Restart all, and the interface notes that some settings apply on the next application start.

Settings Import and Export allow configuration reuse. Keep a known-good export before making extensive changes, and review imported values for the current computer and account setup.

Inspect activation details and understand the available actions.

License Key shows the loaded key, plan/type, activation and expiry details, device binding, recent verification, and hardware information. Use Refresh to request current details. The displayed server response is more useful than static promotional text when determining actual entitlement.

A license backup is not a backup of results, channel files, avatars, or Quick Checker data. Keep it private and use the provider's activation process when restoring access.

Know what is stored and what each removal action affects.

The main workspace stores results in a local SQLite database and maintains local avatar/cache files. Quick Checker has a separate QuickChecker.db, normally under Data with a Data_qc fallback. Use the application's folder-opening controls to locate files in your installed package.

Export my data

Stop filtering and open App Settings → Privacy & Data → Export my data. Choose a ZIP destination. The implemented archive targets the main database, settings, and logs. Do not assume it includes all images, channel-session files, or the separate Quick Checker store. Inspect the ZIP contents before treating it as a complete recovery copy.

Local and external processing

Face/object analysis uses local workers. License verification, WhatsApp requests, avatar fetching, platform orders, and some lookup/tool features involve external services. Therefore, the static UI statement that nothing is uploaded should not be applied to the entire application. Cloud checking explicitly submits a list to a remote service.

Keep account/session files, keys, tokens, result databases, and license backups out of public documentation. This guide contains no real account credentials, phone-number datasets, or profile images from the supplied project.

Use the observed symptom to choose the next check.

Open Logs to filter by AI, Channels, Images, or System and by severity. Search text or use the regular-expression switch. Pause live scrolling, toggle wrapping, copy visible lines, or export a log file. Verbose shows additional uncategorized output. App Settings → Advanced → Open log folder opens the stored logs location.

When asking for support, provide the installed version, module name, exact error, approximate input size, relevant non-sensitive settings, and a redacted log excerpt. Do not send an entire live project folder as a diagnostic shortcut.

Shortcuts are scoped to the active application window.

Quick Checker

Text-entry fields and modal dialogs can change keyboard handling. Use the visible buttons if a shortcut does not apply to the current context. In this manual, Ctrl+K focuses chapter search and Escape clears it.

A compact vocabulary and the source areas behind this manual.

End of manual · PWF 23.5 source snapshot · English edition · September 2026

Who builds Premium WhatsApp Filter, and where to get help.

Premium WhatsApp Filter is built and maintained by PWF Studio, an independent team working on high-throughput WhatsApp tooling and on-device AI. The application validates phone numbers on WhatsApp at scale, downloads profile avatars, and runs face and object recognition on them to classify gender, age band, and visual content. It is aimed at marketers, researchers, and operators who need clean, segmented contact lists.

What the application covers

Support & resources

Scan to open

Point a phone camera at a code to open the destination directly — useful when the manual is being read on paper or on a second screen.

This manual

This guide documents the user-facing workflows of the supplied source snapshot. Live authentication, WhatsApp connectivity, cloud orders, and paid services were not exercised while preparing it. Subscriber numbers, profile photos, hardware identifiers, and local file paths are masked in every screenshot.

© 2026 PWF Studio · All rights reserved.