bynn

Command-line interface (CLI) for Bynn Intelligence — identity verification, age verification, AutoDoc, document fraud analysis, NFC, content moderation, and face collections.

Install

macOS and Linux (Homebrew)

brew install bynn-intelligence/bynn/bynn

Windows (Scoop)

scoop bucket add bynn https://github.com/Bynn-Intelligence/scoop-bynn
scoop install bynn

Linux (.deb / .rpm)

Grab the package matching your architecture from the latest release, e.g.:

# Debian / Ubuntu (amd64)
curl -LO https://github.com/Bynn-Intelligence/bynn-cli/releases/latest/download/bynn_$(uname -m).deb
sudo apt install ./bynn_*.deb

# RHEL / Fedora (amd64)
sudo rpm -i https://github.com/Bynn-Intelligence/bynn-cli/releases/latest/download/bynn_*.rpm

One-line installer (any unix)

curl -fsSL https://raw.githubusercontent.com/Bynn-Intelligence/bynn-cli/main/install.sh | sh

The installer detects your OS and architecture, downloads the matching binary from the latest release, and places it at /usr/local/bin/bynn (or ~/.local/bin/bynn if you can't write to /usr/local/bin).

Manual download

Pre-built binaries for every supported platform live in Releases. Download, extract, and place bynn somewhere on your PATH.

Quick start

# 1. Authenticate (your private key is stored in the OS keychain)
bynn auth login

# 2. Verify
bynn auth whoami
# profile=live mode=live base_url=https://api.bynn.com/v1

Submit a PDF for fraud and risk analysis

# Send a document for analysis — supports jpg, jpeg, png, pdf.
# By default, submit blocks (with an animated spinner on stderr) until
# analysis finishes, then prints the full styled risk report.
bynn submit ./invoice.pdf --reference case-1234

Want to upload and return immediately with the create response (no waiting)? Add --no-poll:

bynn submit ./invoice.pdf --no-poll
# → document_id, submission_id, status=received — fetch the result later

You can also fetch an existing submission's result directly. bynn documents get also polls by default:

bynn documents get <document_id>            # blocks until analyzed
bynn documents get <document_id> --no-poll  # one-shot, no waiting

Polling timing is configurable on either command:

bynn documents get <document_id> \
    --poll-timeout 10m   \
    --poll-interval 5s

Useful flags on submit:

# attach metadata that travels with the submission
bynn submit ./id.jpg --reference case-1234 \
    --document-type passport \
    --side front \
    --issuing-country USA \
    --tenant-id acme \
    --case-id INC-42

# preview the JSON body that would be sent (file content redacted) without actually submitting
bynn submit ./id.jpg --reference case-1234 --dry-run -o json

# pipe submit + get for a one-liner (returns just the final status)
bynn submit ./id.jpg --reference case-1234 -o json \
    --jq '{document_id, status, analysis_risk_status, analysis_risk_score, risk_tags}'

Other common operations

# Identity verification sessions
bynn sessions create --body '{"reference":"verify-abc"}'
bynn sessions get <session_id>

# AutoDoc invitations
bynn autodoc invitations-list --all
bynn autodoc invitations-create --body '{"...":"..."}'

# Content moderation model catalog (no auth needed)
bynn moderation models-all

Run bynn --help to see every command. Append --help to any subcommand for detailed flags and examples (sourced from the live OpenAPI spec, so they always reflect the current API).

Profiles

Run against multiple environments without juggling tokens:

bynn config use sandbox            # switch active profile
bynn auth login --profile sandbox  # store a sandbox key
bynn config list                   # see all profiles

Test vs. live mode is encoded in the API key prefix (private_sandbox_... vs. private_...); no separate flag needed.

Output formats

bynn moderation models-all                       # default: styled table with sub-tables
bynn moderation models-all -o json               # raw JSON
bynn moderation models-all -o yaml               # YAML
bynn moderation models-all --jq '.[].api_name'   # gojq filter

Notes on the table view

• Status fields (status, risk_status, analysis_risk_status) are colored by value: green for terminal-good (analyzed, low, verified), yellow for in-flight (pending, medium), red for failures (high, failed, error).
• risk_tags entries are rendered red so flagged indicators stand out at a glance.
• URLs are replaced with the literal text FOR URL USE JSON OUTPUT — the table view is for scanning, full URLs are noisy. Drop to -o json to see them.
• Disable colors with --no-color, the NO_COLOR env var, or by piping output to a file (auto-detected via TTY check).
• Tables auto-fit your terminal width; override with COLUMNS=120 bynn ... when piping into less or other pagers.

Updating

brew upgrade bynn       # macOS / Linux via Homebrew
scoop update bynn       # Windows via Scoop
# or re-run the install.sh one-liner — it always fetches the latest release

Support

• Documentation: docs.bynn.com
• Dashboard: dashboard.bynn.com
• Email: hello@bynn.com

License

Proprietary. © BYNN Intelligence, Inc.