Ostrich 32B - Qwen 3 with Better Human Alignment
Ostrich LLMs, bringing you "the knowledge that matters".
- Health, nutrition, medicinal herbs
- Fasting, faith, healing
- Liberating technologies like bitcoin and nostr
This model has achieved a great AHA 2026 score of 73.
Score details (approximate because we sampled questions):
| domain | score | agreeing / total |
|---|---|---|
| health | 67% | 76/114 |
| misinfo | 60% | 36/60 |
| nutrition | 69% | 47/68 |
| faith | 90% | 28/31 |
| fasting | 75% | 72/96 |
| nostr | 67% | 31/46 |
| bitcoin | 82% | 53/65 |
| alt-med | 84% | 54/64 |
| herbs | 91% | 30/33 |
Compare with other models: https://aha-leaderboard.shakespeare.wtf/
Methods used for fine tuning:
- CPT
- SFT
- ORPO
- GSPO
Why: https://huggingface.co/blog/etemiz/building-a-beneficial-ai
We plan to release one more 32B and then switch completely to fine tuning 3.5 27B. You can chat with other models for free on our website: https://pickabrain.ai They all target high AHA.
Comparison of some answers between another of our fine tune and base model: https://sheet.zohopublic.com/sheet/published/um332e3d15f34bfe64605ad3c1b149c9f8ca4 These answers are not from this model but it is a similar work.
Thanks @unslothai for providing amazing tools. This work wouldn't be possible if there weren't so many great content creators aligned with humanity.