OpenThai-SystemOne
A Thai-and-English System One model. Qwen3.5-0.8B continued-pretrained on about 5B Thai tokens, its language-model head replaced by a 256-way slot head (slot 255 abstains), trained on 2–3M decision examples and temperature-calibrated per question type.
Version 0.3 scores 74.3% macro across Bespoke's 13 public subsets (Nimble-9B 74.8%, Jev 76.0%); per-subset calibration error ranges from 0.035 to 0.371, with an English median of 0.15. It needs trust_remote_code for its custom architecture, mirrors /v1/systemone, and takes up to 255 options and 2–10 score levels. Latency is 40–48 ms on an H100 with 255 options and 154 ms on an M3 Max. iApp also runs a free hosted endpoint.
What it decides
- choice — picks one option from a set
- score — places the input on an ordered scale
- noul — answers a yes/no question with one calibrated probability
- classify — assigns a category from a fixed taxonomy
- route — sends the input to one of several destinations
At a glance
| Parameters | 0.8B |
| Base model | qwen/qwen3.5-0.8b-base |
| Maker | iApp Technology |
| Released | 2026-09-20 |
| License | apache-2.0 |
| Reported accuracy | 74.3% |
| Reported latency | 40–48 ms on H100 with 255 options; 154 ms on M3 Max |
Get the weights
pip install systemonemodels
systemone pull iapp-technology/openthai-systemone
The files are served from the maker's Hugging Face repository, iapp/OpenThai-SystemOne, and verified against the checksums recorded here.
Read more
This page was opened by System One for iApp Technology, who can claim the organisation and take it over at any time.