交互式3D数学可视化

Deployments

---

Option 2 — Docker image from Hugging Face

Prerequisites: Docker, a Gemini API key.

No clone, no Python setup — run the prebuilt AlgeBench image from the Hugging Face Space registry and pass your own GEMINI_API_KEY:

docker run -it --pull=always -p 7860:7860 --platform=linux/amd64 \
	-e GEMINI_API_KEY="your_key_here" \
	registry.hf.space/ibenian-algebench:latest

Open http://localhost:7860 in your browser.

--pull=always ensures you get the newest published build (Docker otherwise reuses a cached :latest). --platform=linux/amd64 is needed on Apple Silicon and other ARM hosts since the Space image is built for amd64.

To map a different host port, change the left side of -p (e.g. -p 9000:7860) and open that port instead.

Quick Start

There are two ways to run AlgeBench locally. Either way you'll need your own Gemini API key — see Get a Gemini API key.

Option 1 — Run locally from a cloned repo

Prerequisites: uv (recommended) or Python 3.10+, plus a Gemini API key. With uv installed you don't need a preinstalled Python — it provisions the pinned interpreter (.python-version, Python 3.13) for you, on a native CPython build. This matters on Apple Silicon: a bare python3 -m venv often picks an x86 Homebrew Python and runs everything under Rosetta, which roughly halves sympy throughput. With uv present the dev venv is always native arm64. If uv is absent the scripts fall back to python3 -m venv (a warning is printed).

git clone https://github.com/ibenian/algebench
cd algebench
export GEMINI_API_KEY=your_key_here
./algebench

On first run, ./algebench creates a virtual environment and installs the dependencies automatically — no manual pip install needed. Then open http://localhost:8785 in your browser.

To confirm the venv is native (no Rosetta) on Apple Silicon:

.venv/bin/python3 -c "import platform; print(platform.machine())"   # -> arm64

To launch directly into a scene:

./algebench scenes/eigenvalues.json

To update to the latest version of [gemini-live-tools](https://github.com/ibenian/gemini-live-tools) (which includes new voice characters and the voice picker UI):

./algebench --update

This reinstalls gemini-live-tools from GitHub and copies the updated voice-character-selector.js into the app. Not ideal, but simple enough for now.

For all available CLI options including TTS settings:

./algebench --help

Get a Gemini API key

AlgeBench's AI narrator runs on Google's Gemini models, so you need your own free API key from Google AI Studio:

1. Go to aistudio.google.com and sign in with a Google account. 2. Open Get API key (the Get API key button in the left sidebar). 3. Click Create API key. If you don't have a Google Cloud project yet, choose Create API key in new project — AI Studio creates one for you automatically. Otherwise pick an existing project from the list. 4. Copy the generated key. Keep it secret — treat it like a password.

The free tier is enough to try AlgeBench. For higher rate limits, enable billing on the Google Cloud project backing your key.