The November 2026 Abliterated API Index
Silk Compute leads the first edition of the Abliterated API Index at 89.2, 21 points ahead of abliteration.ai. It wins on price, context and the number of 1M-context models.
| Measure | Value | Rank | |
|---|---|---|---|
| Abliterated API Index | 89.2 | 1/6 | |
| Price | $0.405 / 1M | 1/6 | |
| Prompt retention | Never stored | 1/6 | |
| Context window | 1M tokens | 1/6 | |
| 1M-context models | 5 models | 1/6 | |
| API formats | 10 / 10 | 1/6 | |
| Payment and access | 7 / 10 | 2/6 |
The ranking
- Silk Compute, 89.2
- abliteration.ai, 68.2
- Morph, 57.1
- Unfil.AI, 35.4
- Awan LLM, 30.1
- Featherless, 27.5
What decided it
The cheapest 128K-context model costs a blended $0.405 per 1M tokens at Silk Compute and $1.25 at abliteration.ai. Awan LLM and Featherless have no 128K model, so they score no price points.
On retention, 5 of 6 providers say they keep no prompts. Morph keeps them for up to 30 days, and keeps flagged prompts on every tier.
Only three providers serve an abliterated model with 1M tokens of context. Silk Compute serves five, including Kimi K3.1 Pro, which no other ranked provider has.
The full method is on the method page. bestabliterated.com is published by the team behind Silk Compute, one of the providers ranked here.