2026/11/18The November 2026 edition ranks 6 APIs and 8 models
Updated November 18, 2026

Best abliterated models and APIs, ranked

The best abliterated model is Kimi K3.1 Pro Abliterated, and the best abliterated API is Silk Compute at 89.2 out of 100. To run one at home, Qwen 3.8 27B Abliterated fits a single 24 GB GPU.

Top picks

Latest reports

Nov 18, 2026

The November 2026 Abliterated API Index

CostCapabilityPrivacyAbliterated API Index
MeasureValueRank
Abliterated API Index
89.21/6
Price
$0.405 / 1M1/6
Prompt retention
Never stored1/6
Context window
1M tokens1/6
1M-context models
5 models1/6
API formats
10 / 101/6
Payment and access
7 / 102/6
Read from Silk Compute's own pages on November 18, 2026.Full review

Best abliterated models

#ModelParametersContextRuns onLowest price per 1M (in / out)Best for
1Kimi K3.1 Pro AbliteratedMoonshot AI2,780B1MHosted API$3.00 / $8.00Best overallSecurity research
2DeepSeek V4.1 Flash 1017 AbliteratedDeepSeek763B1MHosted API$0.24 / $0.90Best valueCoding agents
3GLM-5.3 AbliteratedZ.ai753B1MHosted API$2.00 / $4.00CodingLong documents
4GLM-5.3 Flash AbliteratedZ.ai320B · 18B active1MHosted API$0.24 / $1.00Lowest input priceHigh volume
5Qwen 3.8 27B AbliteratedAlibaba, abliterated by huihui-ai27.8BVaries24 GB GPU$1.60 / $12.00Best local model24 GB GPU
6Gemma 4 31B HereticGoogle, abliterated with Heretic31.3BVaries24 GB GPUFree weights24 GB GPUWriting
7Qwen 3.6 35B-A3B AbliteratedAlibaba, abliterated by huihui-ai36B · 3B activeVaries24 GB GPUFree weightsFast on a home GPUMixture of experts
8Llama 3.1 8B AbliteratedMeta, abliterated by mlabonne8BVaries8 GB GPU or laptopFree weightsLaptop8 GB GPU
Sizes from Hugging Face cards, and the lowest price among ranked providers.All models

Rankings by job

Best abliterated model for coding agents

Kimi K3.1 Pro Abliterated is the best abliterated model for coding agents, and DeepSeek V4.1 Flash Abliterated is the best value.

See the full ranking
Index scoreBlended price per 1M tokens
Silk ComputeAbliterationMorphUnfil.AIAwan LLMFeatherless
$0.00$0.20$0.40$0.60$0.80$1.00$1.20$1.40
Up and to the left is better, and Awan LLM and Featherless have no 128K model to plot.All providers

Every abliterated API, side by side

Read from each provider's pages on November 18, 2026.

#ProviderBest-value 128K+ modelPrice per 1M (in / out)ContextPrompts storedAnthropic APICryptoIndex
1Silk ComputeDeepSeek V4.1 Flash 1017$0.24 / $0.901MNever storedYesYes89.2
2AbliterationAbliterated Large Flash$1.00 / $2.001MNever storedYesNot listed68.2
3MorphDeepSeek V4.1 Flash Uncensored$0.30 / $1.201MUp to 30 daysYesUSDC57.1
4Unfil.AIHermes 3 405B$1.20 / $1.20128KDeleted within 24 hNot statedYes35.4
5Awan LLMNone$5 to $80 a month8KNever storedNot statedPayPal only30.1
6FeatherlessNoneNone32KNever storedNot statedNo27.5

How the index works

Six measures, with the same weights for every provider.

30 points

Price

20 points

Prompt retention

15 points

Context window

15 points

1M-context models

10 points

API formats

10 points

Payment and access

Questions

Is an abliterated model the same as an uncensored model?

Abliteration is one way to make an uncensored model. It removes the refusal direction without retraining, where an uncensored fine-tune retrains the model.

Do abliterated models lose quality?

A little, on some builds. Large models lose less, which is one reason the 700B-class hosted builds rank highest.

Do abliterated models still refuse?

Sometimes. Some builds still decline in a few categories.

What are abliterated models used for?

Coding agents, security research such as malware analysis, authorized red teaming, and training data for safety classifiers.