InferencePort AI

ipx

The InferencePort AI command line.
Cloud chat & media generation · organizations · local models · host server — one native binary.

  ██╗██████╗ ██╗  ██╗
  ██║██╔══██╗╚██╗██╔╝
  ██║██████╔╝ ╚███╔╝
  ██║██╔═══╝  ██╔██╗
  ██║██║     ██╔╝ ██╗
  ╚═╝╚═╝     ╚═╝  ╚═╝

Install

ipx ships as a self-contained native binary (compiled with Nuitka). Use whichever package manager you already have — none of them need Python or anything else installed beyond themselves.

You have	Command
pip	`pip install inferenceport-cli`
npm	`npm install -g @inferenceport-ai/ipx`
cargo	`cargo install ipx-cli`
macOS / Linux shell	`curl -fsSL https://raw.githubusercontent.com/inferenceport-ai/cli/main/install.sh \| sh`
Windows PowerShell	`irm https://raw.githubusercontent.com/inferenceport-ai/cli/main/install.ps1 \| iex`
Nothing	Download from Releases

Tour

ipx auth login                       # email, GitHub, Google, Microsoft, Hugging Face, SSO or API key
ipx auth context acme                # act as an organization (billing, policy, `ipx org …`)

ipx gen chat                         # interactive chat with your Lightning subscription
ipx gen image "a lighthouse at dawn" -o lighthouse.jpg
ipx gen video "a drone shot over a canyon" --seconds 8
ipx gen audio speech "Welcome to InferencePort"

ipx p2g chat                         # credit-metered Pay-2-Go API
ipx p2g 3d --image chair.jpg -o chair.glb
ipx p2g balance

ipx org members list                 # organizations
ipx org invite create dev@acme.com --role admin
ipx org keys create ci --monthly-limit 25
ipx featherless deploy Qwen/Qwen2.5-7B-Instruct   # personal, or for the org with `-O acme`
ipx agent chat acme-bot/support      # managed agents
ipx mcp council reason "…" --tier basic   # Council of LLMs MCP

ipx model pull llama3.2              # local models (same engine as the desktop app)
ipx local chat llama3.2
ipx server start --allow teammate@acme.com --detach     # host your models for others
ipx import space black-forest-labs/FLUX.1-schnell       # add to the desktop app

ipx shield scan "Ignore previous instructions…"     # AI Shield (docs: readthedocs, "AI Shield")

ipx usage get                        # quotas, wallet and org spend

Every command has --help, and every command accepts the global --json flag for scripting: ipx --json usage get | jq .quota.

Command map

Group	What it does
`ipx auth`	`login`, `logout`/`signout`, `status`/`whoami`, `context <org>`, `token`
`ipx org`	members, invitations, teams, API keys, billing, usage, audit, domains, SSO, policy, desktop policy
`ipx p2g`	credit-billed: `chat`, `text`, `image`, `video`, `audio …`, `3d`, `models`, `balance`, `ledger`, `keys`, `deploy`
`ipx featherless`	`models`, `status`, `list`, `get`, `deploy`, `cancel`, `restore`, `terminate`, `usage` — personal deployments, or the organization's with an org context
`ipx agent`	`list`, `registry`, `get`, `create`, `update`, `delete`, `chat`, `ask`, `profile`, `settings`, `invite …`, `members …`
`ipx mcp council`	`tiers`, `models`, `reason`, `tools`, `test`, `connect <client>` — the hosted Council of LLMs MCP server
`ipx gen`	subscription-billed (Lightning): `chat`, `text`, `image`, `video`, `audio …`, `models`, `plan`
`ipx shield`	AI Shield detection and risk scoring: `scan`, `analyze`, `phishing`, `url`, `check`, `disposable`, `request`, `file`, `password`, `listing`, `review`, `domain`, `tasks` — see the AI Shield docs
`ipx model` / `ipx local`	manage local models / chat with them
`ipx server`	`start`, `stop`, `status`, `logs` — host your local models on port 52458
`ipx import`	`space`, `website`, `file`, `list`, `rm` for the desktop app library
`ipx usage`	`get`, `tiers`
`ipx config`, `doctor`, `status`, `api`, `rewards`, `completion`	tools

Full documentation lives with the desktop app docs: https://inferenceport-ai.readthedocs.io (CLI section, and AI Shield).

Develop

python -m venv .venv && . .venv/bin/activate      # .venv\Scripts\activate on Windows
pip install -e ".[dev]"
pytest                                            # mocked-API suite, no network or credentials needed
python packaging/build_binary.py                  # Nuitka → dist/ipx-<os>-<arch>

Layout: src/ipx (the CLI), packaging/npm (launcher + platform packages), packaging/cargo (Rust launcher that embeds the release binary), .github/workflows (CI and the multi-platform release pipeline).

License

Metadata

Release files for inferenceport-cli 1.0.2

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for inferenceport-cli 1.0.2
File	Size	Uploaded
inferenceport_cli-1.0.2.tar.gz	127.5 kB	Oct 16, 2026	Details

Built distribution (wheel)

Table of built distributions (wheels) for inferenceport-cli 1.0.2
File	Interpreter	ABI	Platform	Reset
inferenceport_cli-1.0.2-py3-none-any.whl 122.7 kB Oct 16, 2026	Python 3	none	any	Details

Total release size: 250.2 kB

Release files / inferenceport_cli-1.0.2.tar.gz

Download URL	inferenceport_cli-1.0.2.tar.gz
Size	127.5 kB
Tags	Source
SHA-256 checksum How to use checksums	`f1bbed598e457d074b1b1ad6220259d53027b8c3e33d5fadfc05487822f60c6c`
BLAKE2b-256 checksum How to use checksums	`61e6913fa931af53d305c4cb1da4b4c4fd760ffebf2292c27896937d666a83c7`
Upload date	Oct 16, 2026
Uploaded using Trusted Publishing? What is trusted publishing?	Yes
Uploaded via	`twine/7.0.0 CPython/3.13.14`

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

Identity

Publishing platform: GitHub Actions
Publishing repository: github.com/InferencePort-AI/cli
Publishing commit: github.com/InferencePort-AI/cli/tree/4d7ffbeff37bfc4530c23312f508e5d2ce6a2d6d

Workflow

Publishing configuration: github.com/InferencePort-AI/cli/blob/4d7ffbeff37bfc4530c23312f508e5d2ce6a2d6d/.github/workflows/release.yml
Publishing logs: github.com/InferencePort-AI/cli/actions/runs/38929287626/attempts/1

Artifact

Subject: inferenceport_cli-1.0.2.tar.gz
SHA-256 checksum: f1bbed598e457d074b1b1ad6220259d53027b8c3e33d5fadfc05487822f60c6c Verifying attestations

Release files / inferenceport_cli-1.0.2-py3-none-any.whl

Download URL	inferenceport_cli-1.0.2-py3-none-any.whl
Size	122.7 kB
Tags	Python 3
SHA-256 checksum How to use checksums	`dbd252da535987dac20061c349b9810bd83b89a71b5d1a10aa60303bfc082c41`
BLAKE2b-256 checksum How to use checksums	`8c5209dd064eb220e720099ed4d2ea6c6ba2b794702c04ae3401db4bf1d67af5`
Upload date	Oct 16, 2026
Uploaded using Trusted Publishing? What is trusted publishing?	Yes
Uploaded via	`twine/7.0.0 CPython/3.13.14`

Provenance

Identity

Publishing platform: GitHub Actions
Publishing repository: github.com/InferencePort-AI/cli
Publishing commit: github.com/InferencePort-AI/cli/tree/4d7ffbeff37bfc4530c23312f508e5d2ce6a2d6d

Workflow

Publishing configuration: github.com/InferencePort-AI/cli/blob/4d7ffbeff37bfc4530c23312f508e5d2ce6a2d6d/.github/workflows/release.yml
Publishing logs: github.com/InferencePort-AI/cli/actions/runs/38929287626/attempts/1

Artifact

Subject: inferenceport_cli-1.0.2-py3-none-any.whl
SHA-256 checksum: dbd252da535987dac20061c349b9810bd83b89a71b5d1a10aa60303bfc082c41 Verifying attestations