# Logits Sigma-1 / Duo decision model by Logits

> Logits Sigma-1 and Duo return option probabilities with open weights. Read their public-suite selection limits, GPU setups and data disclosures.

Updated 2026-11-12 (Australia/Adelaide)

## Quick answers

### What is Logits Sigma-1 / Duo by Logits?

[Logits Sigma-1 / Duo](<https://decisionmodels.ai/models/logits-sigma>): Two text decision packages.

[Logits Sigma-1 / Duo facts](<https://decisionmodels.ai/models/logits-sigma>)

[Source](<https://huggingface.co/logitshq/Logits-Sigma-1>)

### Who makes Logits Sigma-1 / Duo?

[Logits Sigma-1 / Duo](<https://decisionmodels.ai/models/logits-sigma>): Logits

[Logits Sigma-1 / Duo facts](<https://decisionmodels.ai/models/logits-sigma>)

[Source](<https://huggingface.co/logitshq/Logits-Sigma-1>)

### Are Logits Sigma-1 / Duo's weights open?

[Logits Sigma-1 / Duo](<https://decisionmodels.ai/models/logits-sigma>): Public weights and Decision Index engines; one 96 GB GPU in the reported setups Licence: Apache-2.0 code and weights; training sources include share-alike, copyleft-labelled and possibly non-commercial data

[Logits Sigma-1 / Duo facts](<https://decisionmodels.ai/models/logits-sigma>)

[Source](<https://huggingface.co/logitshq/Logits-Sigma-1>)

By Logits · Weights upload: 2026-11-07 · OPEN

Two text decision packages. Sigma-1 merges a LoRA into Gemma 4 31B and reads option-code logits. Duo averages option probabilities from a modified [KnowLine](<https://decisionmodels.ai/models/knowline>) 31B and d3-ultra, both in FP8. Above 50,000 request characters, Duo uses its first member alone.

## Facts

- Inputs: Text state and typed choice or yes/no questions
- Licence: Apache-2.0 code and weights; training sources include share-alike, copyleft-labelled and possibly non-commercial data
- Access: Public weights and Decision Index engines; one 96 GB GPU in the reported setups
- Limits: Sigma-1 refuses prompts above 131,072 tokens or menus above 255 options

## Claims and sources

- **VENDOR-REPORTED**: Logits reports a public Decision Index 0.3 score of 67.1 for Sigma-1 over 140,178 scoreable requests. Model selection used a subset of the public suite, making these scores partly in-sample. No private Full score is supplied. [source](<https://huggingface.co/logitshq/Logits-Sigma-1>)
- **VENDOR-REPORTED**: Logits reports a public Decision Index 0.3 score of 72.93 for Duo over 140,178 scoreable requests. Model selection used a subset of the public suite, making these scores partly in-sample. No private Full score is supplied. [source](<https://huggingface.co/logitshq/Logits-Sigma-1-Duo>)
- **VENDOR-REPORTED**: On one RTX PRO 6000, the maker reports median/mean request times of 122/356 ms for Sigma-1. [source](<https://huggingface.co/logitshq/Logits-Sigma-1>)
- **VENDOR-REPORTED**: On one RTX PRO 6000, the maker reports median/mean request times of 293/859 ms for Duo, with GPU memory peaking at 87,891 MiB. Its published training disclosure includes source-licence restrictions that the package licence does not settle. [source](<https://huggingface.co/logitshq/Logits-Sigma-1-Duo>)

## Links

- [Source: Sigma-1 card and weights](<https://huggingface.co/logitshq/Logits-Sigma-1>)
- [Source: Duo card and weights](<https://huggingface.co/logitshq/Logits-Sigma-1-Duo>)
- [Source: Duo source-data disclosure](<https://huggingface.co/logitshq/Logits-Sigma-1-Duo/blob/main/DATA_SOURCES.md>)

## Related projects

No related projects are listed in the registry.