# Remek Kinas: basal

> Polish and English typed-decision models fine-tuned from SpeakLeash's Bielik. A state goes in; choice, yes/no (noul) and score questions come back as a calibrated probability per option, read from the option-letter logits in one pass per option order, with no generated text.

- Page: https://systemonemodels.ai/remek-kinas/basal
- API: https://api.systemonemodels.ai/v1/models/remek-kinas/basal
- Download: `pip install systemonemodels && systemone pull remek-kinas/basal`

## Facts

| | |
|---|---|
| Maker | Remek Kinas (https://systemonemodels.ai/remek-kinas) |
| Decides | choice, score, noul, classify, route |
| Architecture | basal |
| Base model | speakleash/bielik-4.5b-v3.0-instruct |
| Parameters | 4.5B |
| Licence | apache-2.0 |
| Availability | Open weights |
| Released | 2026-10-05 |
| Latest version | 1.5.0 |

## Reported evaluation

Suite: Werdykt v1 (the maker's hidden Polish-English typed-decision benchmark, 10 categories x 2 languages, macro accuracy), 2026-10-03. Numbers are the publisher's own.

- Decision accuracy: 72.1%

- Median latency: 33.8 ms

## Model card

<!-- generated by scripts/seed_catalog.py; edit content/models/catalog.yaml -->

# basal-1.5

Polish and English typed-decision models fine-tuned from SpeakLeash's Bielik. A state goes in; choice, yes/no (noul) and score questions come back as a calibrated probability per option, read from the option-letter logits in one pass per option order, with no generated text.

Polish first, English second. The served answer averages two passes, one per option order, then applies a per-type temperature; the confidence thresholds are validated only on some prompt types. The bundled engine (rkinas/basal) serves /v1/systemone and adds experimental multi, act and evidence-span types; it runs on PyTorch, vLLM, SGLang, MLX, Ollama and llama.cpp. Up to 10 options per question. Siblings, all Apache-2.0: basal-1.5-max (11B, the maker's most accurate), basal-1.5-mini (1.5B) and basal-1.0 (26 September, with a technical report on Zenodo). The maker's own numbers: 0.931 on a sealed Polish test, 0.688 on ten public English tasks; Werdykt is the maker's own hidden benchmark, with only samples public.

## 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 probability
- **classify** — assigns a category from a fixed taxonomy
- **route** — sends the input to one of several destinations

## At a glance

| | |
|---|---|
| Parameters | 4.5B |
| Base model | `speakleash/Bielik-4.5B-v3.0-Instruct` |
| Maker | Remek Kinas |
| Released | 2026-10-05 |
| License | apache-2.0 |
| Reported accuracy | 72.1% |
| Reported latency | 33.8 ms median per decision with SGLang on one H100 |

## Get the weights

```bash
pip install systemonemodels
systemone pull remek-kinas/basal
```

The files are served from the maker's Hugging Face repository, [`Remek/basal-1.5-4.5B`](https://huggingface.co/Remek/basal-1.5-4.5B), and verified against the checksums recorded here.

## Read more

- [Model card](https://huggingface.co/Remek/basal-1.5-4.5B)
- [Inference engine](https://github.com/rkinas/basal)
- [Website and Werdykt leaderboard](https://basal.si5.pl/)
- [basal-1.5-max](https://huggingface.co/Remek/basal-1.5-max)
- [basal-1.0 technical report](https://doi.org/10.5281/zenodo.23022986)

---

*This page was opened by System One for Remek Kinas, who can claim the organisation and take it over at any time.*

---

From System One Models — https://systemonemodels.ai/ · every System One model: https://systemonemodels.ai/system-one-models