ai999x-encoder
Ask in your own words — find the right formula across every AI999X model.
About this model
ai999x-encoder reads a question written in plain English and finds which AI999X capability it is about — a formula, a procedure or a law — across every AI999X model at once. At 4.9 MB it runs offline on any CPU and right in the browser: it powers search in every AI999X Python package and the capability orb.
Install
pip install physics-ai999x[search] hf download AI999X/ai999x-encoder --local-dir ai999x-encoder
Use
# pip install "physics-ai999x[search]"
import physics_ai999x as physics
physics.search("How long does a pendulum take to swing back and forth?", k=3) Topics
Versions
| Version | Released |
|---|---|
| 1.23.2 | 2026-10-06 |
| 1.23.1 | 2026-09-29 |
| 1.23.0 | 2026-09-18 |
| 1.23.0rc1 | 2026-09-02 |
| 1.22.0 | 2026-06-15 |
| 1.22.0rc2 | 2026-06-11 |
| 1.22.0rc1 | 2026-06-02 |
| 1.21.0 | 2026-03-27 |
| 1.21.0rc4 | 2026-03-23 |
| 1.21.0rc3 | 2026-03-15 |
| 1.21.0rc2 | 2026-03-11 |
| 1.21.0rc1 | 2026-03-06 |
Files
9 files, 43.3 MB in total. The largest:
| File | Size |
|---|---|
| onnx/encoder.onnx | 18.9 MB |
| model.safetensors | 18.9 MB |
| onnx/encoder.int8.onnx | 4.9 MB |
| capability_index.json | 403.3 KB |
| tokenizer.json | 176.4 KB |
| LICENSE | 11.4 KB |
| README.md | 4.8 KB |
| evaluation/report.json | 2.5 KB |
| config.json | 511 B |
More natural language processing models
More from AI999X