# MoDe S / M / L decision model by newrexlabs

> MoDe has three sparse decision models for English text and JSON. Read their noncommercial licence, public-suite scores and reported GPU timing limits.

Updated 2026-11-12 (Australia/Adelaide)

## Quick answers

### What is MoDe S / M / L by newrexlabs?

[MoDe S / M / L](<https://decisionmodels.ai/models/mode>): Three sparse MobileMoE decision fine-tunes for English text or JSON.

[MoDe S / M / L facts](<https://decisionmodels.ai/models/mode>)

[Source](<https://huggingface.co/newrexlabs/MoDe-S>)

### Who makes MoDe S / M / L?

[MoDe S / M / L](<https://decisionmodels.ai/models/mode>): newrexlabs

[MoDe S / M / L facts](<https://decisionmodels.ai/models/mode>)

[Source](<https://huggingface.co/newrexlabs/MoDe-S>)

### Are MoDe S / M / L's weights open?

[MoDe S / M / L](<https://decisionmodels.ai/models/mode>): Public weights and MoDe inference package; custom model code requires trust\_remote\_code=True Licence: Weights: FAIR Noncommercial Research License and Acceptable Use Policy; inference code: Apache-2.0

[MoDe S / M / L facts](<https://decisionmodels.ai/models/mode>)

[Source](<https://huggingface.co/newrexlabs/MoDe-S>)

By newrexlabs · Weights upload: 2026-11-07 · NONCOMMERCIAL

Three sparse MobileMoE decision fine-tunes for English text or JSON. They score supplied options in one forward pass, with a pointer head and temperature scaling by question type. Weights use a noncommercial research licence; inference code is Apache-2.0. Training code and data are not released.

## Facts

- Size: S: 1.27B total / 0.28B active; M: 2.82B / 0.53B; L: 5.33B / 0.93B
- Inputs: English text or JSON; choice, noul and score
- Context: 8,192 tokens per pass; split longer requests by question, reject a single question that does not fit
- Licence: Weights: FAIR Noncommercial Research License and Acceptable Use Policy; inference code: Apache-2.0
- Access: Public weights and MoDe inference package; custom model code requires trust\_remote\_code=True

## Claims and sources

- **VENDOR-REPORTED**: The maker reports public Decision Index 0.3 scores of 42.02, 47.10 and 51.36 for S, M and L on 140,178 requests. Four overlength requests count as wrong. These submissions have not been verified by the benchmark maintainers. [source](<https://huggingface.co/newrexlabs/MoDe-S>)
- **VENDOR-REPORTED**: The maker reports medians of 11.0, 17.7 and 25.0 ms for S, M and L on a half-card RTX PRO 6000 MIG partition. Each is the median of five run medians over 700 requests after ten warm-ups, using bf16 and each model’s own engine. These are not full-card leaderboard timings. [source](<https://huggingface.co/newrexlabs/MoDe-S>)

## Links

- [Source: Small card and weights](<https://huggingface.co/newrexlabs/MoDe-S>)
- [Source: Medium card and weights](<https://huggingface.co/newrexlabs/MoDe-M>)
- [Source: Large card and weights](<https://huggingface.co/newrexlabs/MoDe-L>)
- [Source: Inference code](<https://github.com/newrexlabs/MoDe>)
- [Source: Licence](<https://huggingface.co/newrexlabs/MoDe-S/blob/main/LICENSE>)

## Related projects

No related projects are listed in the registry.