Listed in Models
Nous Research's open fine-tunes, from 14B up to 405B. Pricing, the evidence and the alternatives, checked Oct 2026.
Open weights on Hugging Face, Apache-2.0 (4.3-36B, 14B); Llama-based sizes follow Llama license. 14B, 36B, 70B, 405B. 14B and 36B run locally; 70B/405B need a server.
Nous Research's open fine-tunes, from 14B up to 405B.
Hermes 4: Nous Research's open fine-tunes, from 14B up to 405B. Best for steerable, lightly filtered local chat and role-play on your own hardware. Watch out for: far behind frontier models on benchmarks (aa index 8 for the 70b); newest is from dec 2025. Price: Free to download; hosted price via Nous Portal not verified (no free tier, checked Oct 2026).
| Free tier | Free weights to self-host |
|---|---|
| Job in a stack | AI model, open-weights model |
| Cheapest paid plan | API price: unverified |
| Price checked | 2026-10-15, unverified |
Whole scores from 1 to 5, mapped from the source shown. “Not rated” means we found no public benchmark for that job; “not supported” means the model can’t do it. How the bars are scored.
| Plan | Price | Billing | Key limits |
|---|---|---|---|
| We couldn’t read plan prices from the official page, so these are unverified. | |||
From portal.nousresearch.com/, checked 2026-10-15, unverified. Some prices could not be read from the official page, so confirm before buying. Prices are before tax and vendors change them often.
| Type | Source | Result | Date |
|---|---|---|---|
| benchmark | Artificial Analysis Intelligence Index | 8 (Hermes 4 70B (AA marks it * estimated)) | 2026-10-15 |
| adoption | Hugging Face downloads | 1,641 downloads last month; license Apache-2.0 | 2026-10-15 |
| adoption | Nous Research releases | Hermes-4.3-Seed-36B released 2025-12-03 | 2026-10-15 |
Our scores: quality 2/5, ease 3/5, value 2.5/5. Editorial start 2/3; benchmark adjustment +0, rating adjustment +0. How we rank.
No free tier. Cheapest paid plan: Free to download; hosted price via Nous Portal not verified (checked Oct 2026).
Steerable, lightly filtered local chat and role-play on your own hardware.
Far behind frontier models on benchmarks (AA index 8 for the 70B); newest is from Dec 2025.