---
title: "Best Abliterated Models in 2026: GLM 5.3, DeepSeek V4.1 Flash and Qwen 3.8 Compared"
description: "The best abliterated LLMs in 2026: GLM 5.3 for coding, DeepSeek V4.1 Flash for the fewest refusals, compared on benchmarks, refusal rates and API price."
canonical: "https://allmodels.io/guides/best-abliterated-models"
markdown: "https://allmodels.io/guides/best-abliterated-models.md"
author: "AllModels"
datePublished: "2026-11-03"
dateModified: "2026-11-04"
---

# Best abliterated models in 2026

How they score on independent benchmarks and how completely their refusals have been removed.

![The Qwen, GLM by Z.ai and DeepSeek logos](https://allmodels.io/assets/guides/abliterated-models-2026.svg)

*The best abliterated models of 2026 are built on three open model families: Qwen from Alibaba, GLM from Z.ai and DeepSeek.*

## What is abliteration?

An abliterated model is an open-weight LLM whose refusal behaviour has been removed by editing its weights, with no retraining. Research in 2024 showed that refusals in chat models are controlled by a single direction in the model's internal activations, and the first abliterated Llama 3 models removed that direction from the weights. The name is a blend of "ablate" and "obliterate".

Abliteration techniques have changed a lot since the start of 2026, and the field is still moving fast. Removing one refusal direction has given way to removing several: Heretic's arbitrary-rank ablation, a rank-2 refusal subspace for DeepSeek V4.1 Flash with separate directions for chat and reasoning, and edits tuned for each reasoning-effort level. How well it works now depends a lot on the model. A method that removes almost every refusal from one model can leave many behind in another.

The leading abliterated builds come from a few publishers on Hugging Face. **dealignai** publishes weight-level uncensored builds of GLM 5.3, GLM 5.3 Flash and DeepSeek V4.1 Flash 1017, each with published HarmBench refusal and MMLU results. **orcarouter** publishes uncensored Qwen 3.8, GLM 5.3 Flash and DeepSeek V4.1 Flash builds. **Audn AI** publishes Penclaw, a line of abliterated frontier models that includes GLM 5.3. **p-e-w** created Heretic, the open-source abliteration tool behind thousands of community models, and also publishes Heretic builds of Qwen models.

### From Qwen to GLM 5.3

At the start of 2026, an abliterated Qwen model was the usual answer to which abliterated model is best. That changed when Z.ai released GLM 5.3 in August. GLM 5.3 has been the most important open model of the year for software development, and it still leads DeepSeek V4.1 Flash on agentic coding, even though DeepSeek released V4.1 Flash after it and updated it in October.

GLM 5.3 comes in two variants with different architectures. **GLM 5.3** is a 753B mixture-of-experts model that takes text only. **GLM 5.3 Flash** is a 320B model with 18B active parameters, and it is multimodal: it accepts images as well as text. The full model costs more per token, but as of November 2026 it is the best open model for software development, and its abliterated version is the best abliterated model for coding.

### How GLM 5.3, DeepSeek V4.1 Flash and Qwen 3.8 rank

Artificial Analysis runs the same evals on every major model. On version 5 of its Intelligence Index, DeepSeek V4.1 Flash 1017 and GLM 5.3 are the two highest-scoring open-weight models, at 33 and 32. GLM 5.3 Flash scores 28 and Qwen 3.8 27B scores 20, the second best among small open models.

On coding, GLM 5.3 is ahead. It scores 41.9% on Terminal-Bench 4.0, the agentic coding eval, against 33.8% for DeepSeek V4.1 Flash, and 59.0% on SciCode against 55.9%. It also leads on CritPt, a physics research eval, and hallucinates far less: it gives a wrong answer instead of admitting uncertainty on 39% of the knowledge questions it misses, against 64% for DeepSeek. DeepSeek V4.1 Flash leads on general reasoning and knowledge, and costs about an eighth as much per task.

#### Artificial Analysis rankings: intelligence, coding and cost

**Intelligence Index** (Higher is better)

1. DeepSeek V4.1 Flash: 32.6
2. GLM 5.3: 31.7
3. GLM 5.3 Flash: 27.8
4. Qwen 3.8 27B: 20.1

**Agentic coding (Terminal-Bench 4.0)** (Higher is better)

1. GLM 5.3: 41.9%
2. DeepSeek V4.1 Flash: 33.8%
3. GLM 5.3 Flash: 32.8%
4. Qwen 3.8 27B: 5.6%

**Cost per Intelligence Index task** (Lower is better)

1. GLM 5.3 Flash: $0.58
2. DeepSeek V4.1 Flash: $0.60
3. Qwen 3.8 27B: $3.53
4. GLM 5.3: $5.01

*Source: Artificial Analysis, November 2026. Original models, not the abliterated builds.*

#### GLM 5.3, DeepSeek V4.1 Flash and Qwen 3.8 on Artificial Analysis evals

| Model | Intelligence Index | Terminal-Bench 4.0 | SciCode | CritPt | Hallucination rate | FrontierCode 2.0 | HLE-Diamond | Cost per task |
| --- | --- | --- | --- | --- | --- | --- | --- | --- |
| GLM 5.3 | 31.7 | **41.9%** | **59.0%** | **33.5%** | 38.9% | 21.6% | 17.9% | $5.01 |
| GLM 5.3 Flash | 27.8 | 32.8% | 51.6% | 24.5% | 38.4% | 17.6% | 18.2% | **$0.58** |
| DeepSeek V4.1 Flash 1017 | **32.6** | 33.8% | 55.9% | 25.0% | 64.1% | **28.5%** | **25.3%** | $0.60 |
| Qwen 3.8 27B | 20.1 | 5.6% | 46.6% | 4.0% | **28.6%** | 14.9% | 15.3% | $3.53 |

*Source: Artificial Analysis, November 2026.*

### GLM 5.3 still refuses some prompts

Abliteration has worked less well on GLM 5.3 than on other models. Most researchers working on it have not been able to remove its refusals completely, and classic abliteration leaves more refusals behind than it does on Qwen or DeepSeek. Even the best public release comes with caveats: dealignai's GLM 5.3 build fully answers about 80% of HarmBench prompts and still refuses 12–13% of them.

### Which abliterated model refuses most

dealignai tests each of its builds on the same 320 HarmBench prompts, so its results show how the models compare after abliteration. GLM 5.3 refuses the most: it still refuses 12–13% of the prompts, with soft refusals or redirects rather than outright refusals. GLM 5.3 Flash and DeepSeek V4.1 Flash 1017 answer all 320 at their default reasoning effort. GLM 5.3 Flash keeps some refusals at low reasoning effort, by design. Both GLM builds score slightly higher on MMLU than the original models, while DeepSeek loses about 4 points.

#### Refusals after abliteration: dealignai's GLM 5.3, GLM 5.3 Flash and DeepSeek V4.1 Flash builds

| Model | Answered in full | Still refused | Reasoning effort tested | MMLU, original → abliterated |
| --- | --- | --- | --- | --- |
| GLM 5.3 | 79–82% | 12–13% | Off and max | 85.6% → 87.4% |
| GLM 5.3 Flash | **100%** | **0%** | Off and max | 86.7% → 87.3% |
| DeepSeek V4.1 Flash 1017 | **100%** | **0%** | Max | 87.6% → 83.8% |

*Source: dealignai model cards on Hugging Face, November 2026. 320 HarmBench prompts.*

- **Still refused** includes soft refusals and redirects: answers that decline politely, hedge or steer to a safe alternative.

### Use a router

Running these models locally generally means quantizing them to fit the hardware, and quantization substantially degrades the quality of their outputs. That makes a local build unsuitable for serious security-related work.

Mission-critical workloads also need near-perfect uptime, and many providers of these models have significant uptime and latency problems. That is why a router is the best approach. The AllModels [Abliterated Model Router](https://allmodels.io/llm/abliterated) serves GLM 5.3, GLM 5.3 Flash and DeepSeek V4.1 Flash Abliterated from two to five providers each. If one provider cannot serve a request, AllModels sends it to the next before output starts, so long agentic sessions keep running.

## Abliterated models on AllModels compared

Every model here is available through the AllModels API. Prices are per million input / output tokens, from the cheapest provider.

Abliterated models on AllModels. Checked November 2026.

| Product | Best for | Input | Refusal behavior | Providers | Price / 1M tokens |
| --- | --- | --- | --- | --- | --- |
| [GLM 5.3 Abliterated](https://allmodels.io/llm/models/z-ai/glm-5.3-abliterated) | The most difficult work, including coding and agents | Text | 12–13% still refused | 3 | From $1.06 / $2.11 |
| GLM 5.3 Flash Abliterated | Cheap GLM with image input | Text and images | Fully removed | 2 | From $0.13 / $0.53 |
| DeepSeek V4.1 Flash Abliterated | Best quality for the cost, and the fewest refusals | Text and images | Fully removed | 5 | From $0.15 / $0.53 |
| Qwen 3.8 27B Abliterated | Small, fast general use | Text | Fully removed | 1 | $2.11 / $8.44 |


### Each model in brief

#### GLM 5.3 Abliterated

The best abliterated model for software development as of November 2026, and the leader on agentic coding. It takes text only and costs more per token than the Flash models. Its refusals are not fully removed, so some prompts still get a hedged or redirected answer.

**Pick it if:** you want the best abliterated model for coding and can live with occasional refusals.

#### GLM 5.3 Flash Abliterated

The smaller GLM 5.3 variant, with 18B active parameters. It accepts images as well as text and is the cheapest model here. dealignai's build answers every HarmBench prompt at its default reasoning effort, but keeps some refusals at low effort.

**Pick it if:** you want a cheaper abliterated GLM, or you need image input.

#### DeepSeek V4.1 Flash Abliterated

The most permissive model here, the top open-weight model on the Artificial Analysis Intelligence Index, and one of the cheapest to call. Published abliterated builds take its refusals to effectively zero. AllModels serves it from five providers.

**Pick it if:** GLM 5.3's leftover refusals get in the way of your work, or cost matters most.

#### Qwen 3.8 27B Abliterated

The most downloaded abliterated model family on Hugging Face, and the usual pick before GLM 5.3 arrived. It is the second best small open model on the Intelligence Index and has the lowest hallucination rate here, but it trails the larger models on coding.

**Pick it if:** you want a small general-purpose model with near-zero refusals.

## The best abliterated models in 2026: GLM 5.3 for difficult work, DeepSeek V4.1 Flash for value

For the most difficult work, **GLM 5.3 Abliterated** is the best, the full model rather than Flash. It leads the open models on agentic coding as of November 2026, and on AllModels it is one API call with the OpenAI SDK you already use.

If you run into refusals, **DeepSeek V4.1 Flash Abliterated** is the best option for quality against cost. It answers every HarmBench prompt in dealignai's tests, scores slightly higher than GLM 5.3 on the Intelligence Index, and costs a fraction as much: from $0.15 / $0.53 per million input / output tokens on AllModels, against $1.06 / $2.11 for GLM 5.3. If you mostly need speed and low cost, the standard `deepseek/deepseek-v4.1-flash-ultrafast` model refuses little and may be all you need.

- **Use a router.** Providers of abliterated models often have uptime, speed and reliability problems. On AllModels, GLM 5.3 Abliterated has three providers, GLM 5.3 Flash Abliterated two and DeepSeek V4.1 Flash Abliterated five. If one cannot serve a request, AllModels tries the next before output starts, so agentic sessions keep running.
- **Pin a provider when you need to.** Choose, order or exclude providers on each request, and read the `provider` field in every response.
- **No prompts stored.** AllModels does not store or log prompts and outputs, or use them for training.
- **Pay as you go.** Provider price plus a 5.5% platform fee, with free credits when you sign up.

Browse every [abliterated model on AllModels](https://allmodels.io/llm/abliterated) or start with free credits.

## Call an abliterated model with the OpenAI SDK

Point the OpenAI SDK, or any HTTP client, at AllModels and use an abliterated model ID. Pick a client, an API and a language; highlighted lines are the AllModels base URL, the model and the provider preference.

### OpenAI SDK, Chat Completions — JavaScript · openai

Client: OpenAI SDK · Mode: Chat Completions · Language: JS

Set the OpenAI SDK base URL to `https://api.allmodels.io/oai` and call an abliterated model by its AllModels ID. [Read the LLM quickstart](https://allmodels.io/llm/docs).

```javascript
import OpenAI from "openai";

const client = new OpenAI({
  apiKey: process.env.ALLMODELS_API_KEY,
  baseURL: "https://api.allmodels.io/oai",
});

const completion = await client.chat.completions.create({
  model: "deepseek/deepseek-v4.1-flash-abliterated",
  messages: [{ role: "user", content: "Explain abliteration in one sentence." }],
});

console.log(completion.choices[0].message.content);
```

### OpenAI SDK, Chat Completions — Python · openai

Client: OpenAI SDK · Mode: Chat Completions · Language: Python

Set the OpenAI SDK base URL to `https://api.allmodels.io/oai` and call an abliterated model by its AllModels ID. [Read the LLM quickstart](https://allmodels.io/llm/docs).

```python
import os
from openai import OpenAI

client = OpenAI(
    api_key=os.environ["ALLMODELS_API_KEY"],
    base_url="https://api.allmodels.io/oai",
)

completion = client.chat.completions.create(
    model="deepseek/deepseek-v4.1-flash-abliterated",
    messages=[{"role": "user", "content": "Explain abliteration in one sentence."}],
)

print(completion.choices[0].message.content)
```

### OpenAI SDK, Responses — JavaScript · openai

Client: OpenAI SDK · Mode: Responses · Language: JS

Use the Responses API and list the providers you prefer. If the first one cannot serve the request, AllModels tries the next before output starts. [Read the provider routing docs](https://allmodels.io/llm/docs/providers).

```javascript
import OpenAI from "openai";

const client = new OpenAI({
  apiKey: process.env.ALLMODELS_API_KEY,
  baseURL: "https://api.allmodels.io/oai",
});

const response = await client.responses.create({
  model: "deepseek/deepseek-v4.1-flash-abliterated",
  input: "Explain abliteration in one sentence.",
  provider: { order: ["hopper", "refuseless"] },
});

console.log(response.output_text);
```

### OpenAI SDK, Responses — Python · openai

Client: OpenAI SDK · Mode: Responses · Language: Python

Use the Responses API and pass your provider preference in `extra_body`. If the first provider cannot serve the request, AllModels tries the next before output starts. [Read the provider routing docs](https://allmodels.io/llm/docs/providers).

```python
import os
from openai import OpenAI

client = OpenAI(
    api_key=os.environ["ALLMODELS_API_KEY"],
    base_url="https://api.allmodels.io/oai",
)

response = client.responses.create(
    model="deepseek/deepseek-v4.1-flash-abliterated",
    input="Explain abliteration in one sentence.",
    extra_body={"provider": {"order": ["hopper", "refuseless"]}},
)

print(response.output_text)
```

### HTTP, Chat Completions — JavaScript · Node.js · fetch

Client: HTTP · Mode: Chat Completions · Language: JS

POST to `https://api.allmodels.io/oai/chat/completions`. The `provider` field in the response tells you which provider served it. [Read the LLM quickstart](https://allmodels.io/llm/docs).

```javascript
const res = await fetch("https://api.allmodels.io/oai/chat/completions", {
  method: "POST",
  headers: {
    Authorization: `Bearer ${process.env.ALLMODELS_API_KEY}`,
    "Content-Type": "application/json",
  },
  body: JSON.stringify({
    model: "deepseek/deepseek-v4.1-flash-abliterated",
    messages: [{ role: "user", content: "Explain abliteration in one sentence." }],
  }),
});

const completion = await res.json();
console.log(completion.provider, completion.choices[0].message.content);
```

### HTTP, Chat Completions — Python · requests

Client: HTTP · Mode: Chat Completions · Language: Python

POST to `https://api.allmodels.io/oai/chat/completions`. The `provider` field in the response tells you which provider served it. [Read the LLM quickstart](https://allmodels.io/llm/docs).

```python
import os
import requests

res = requests.post(
    "https://api.allmodels.io/oai/chat/completions",
    headers={"Authorization": f"Bearer {os.environ['ALLMODELS_API_KEY']}"},
    json={
        "model": "deepseek/deepseek-v4.1-flash-abliterated",
        "messages": [{"role": "user", "content": "Explain abliteration in one sentence."}],
    },
)

completion = res.json()
print(completion["provider"], completion["choices"][0]["message"]["content"])
```

### HTTP, Responses — JavaScript · Node.js · fetch

Client: HTTP · Mode: Responses · Language: JS

POST to `https://api.allmodels.io/oai/responses` with a provider preference, then read the message from `output`. [Read the provider routing docs](https://allmodels.io/llm/docs/providers).

```javascript
const res = await fetch("https://api.allmodels.io/oai/responses", {
  method: "POST",
  headers: {
    Authorization: `Bearer ${process.env.ALLMODELS_API_KEY}`,
    "Content-Type": "application/json",
  },
  body: JSON.stringify({
    model: "deepseek/deepseek-v4.1-flash-abliterated",
    input: "Explain abliteration in one sentence.",
    provider: { order: ["hopper", "refuseless"] },
  }),
});

const response = await res.json();
const message = response.output.find((item) => item.type === "message");
console.log(response.provider, message.content[0].text);
```

### HTTP, Responses — Python · requests

Client: HTTP · Mode: Responses · Language: Python

POST to `https://api.allmodels.io/oai/responses` with a provider preference, then read the message from `output`. [Read the provider routing docs](https://allmodels.io/llm/docs/providers).

```python
import os
import requests

res = requests.post(
    "https://api.allmodels.io/oai/responses",
    headers={"Authorization": f"Bearer {os.environ['ALLMODELS_API_KEY']}"},
    json={
        "model": "deepseek/deepseek-v4.1-flash-abliterated",
        "input": "Explain abliteration in one sentence.",
        "provider": {"order": ["hopper", "refuseless"]},
    },
)

response = res.json()
message = next(item for item in response["output"] if item["type"] == "message")
print(response["provider"], message["content"][0]["text"])
```

## Keep the SDK you already use

### OpenAI Chat Completions

Endpoint: `https://api.allmodels.io/oai/chat/completions`

Point the official OpenAI SDK at `https://api.allmodels.io/oai` and keep calling `chat.completions.create`, now with abliterated model IDs. Responses come back as JSON or stream as server-sent events.

- [Read the LLM quickstart](https://allmodels.io/llm/docs)

### OpenAI Responses

Endpoint: `https://api.allmodels.io/oai/responses`

Use `responses.create` with the same client and key. The Responses API is stateless, so send the full conversation with each request.

- [Read the LLM quickstart](https://allmodels.io/llm/docs)

## Abliterated model questions

### What is the best abliterated model in 2026?

For the most difficult work, GLM 5.3 Abliterated, the full model rather than Flash. Its refusals are not completely removed, so if they get in your way use DeepSeek V4.1 Flash Abliterated, the best option for quality against cost. Both are available through the [AllModels API](https://allmodels.io/llm/abliterated).

### Is GLM 5.3 better than DeepSeek V4.1 Flash?

For coding, yes. On Artificial Analysis, GLM 5.3 scores 41.9% on Terminal-Bench 4.0 against 33.8%, and 59.0% on SciCode against 55.9%, and it hallucinates less. DeepSeek V4.1 Flash scores slightly higher on the overall Intelligence Index and costs about an eighth as much per task.

### Why does abliterated GLM 5.3 still refuse some prompts?

Classic abliteration has worked less well on GLM 5.3 and GLM 5.3 Flash than on other models. Most public builds leave some refusals, hedges or redirects behind. Abliteration removes DeepSeek V4.1 Flash's refusals far more completely.

### Do I need an abliterated version of DeepSeek V4.1 Flash?

Often not. DeepSeek V4.1 Flash refuses little outside clearly harmful requests, and the standard `deepseek/deepseek-v4.1-flash-ultrafast` model on AllModels is very fast and cheap. Use `deepseek/deepseek-v4.1-flash-abliterated` when you need refusals at effectively zero.

### What is the difference between GLM 5.3 and GLM 5.3 Flash?

GLM 5.3 is the full 753B model. It takes text only, costs more and is the stronger model for software development. GLM 5.3 Flash has 320B parameters with 18B active, accepts images as well as text, and is much cheaper.

### What does abliterated mean?

An abliterated model has had its refusal direction removed from its weights, so it answers prompts it would normally decline. It is a weight edit, not a fine-tune, and the rest of the model stays close to the original.

## Use an abliterated model in three steps

1. **Choose a model.** Pick one from the [abliterated model catalogue](https://allmodels.io/llm/abliterated), such as `z-ai/glm-5.3-abliterated` for coding or `deepseek/deepseek-v4.1-flash-abliterated` for the fewest refusals.
2. **Change the base URL.** Point the OpenAI SDK at `https://api.allmodels.io/oai` and use your AllModels API key.
3. **Send your prompt.** Use Chat Completions or Responses. Add a provider preference when you want a specific route.

## Every abliterated model, one API.

Start with free credits and call frontier abliterated models without running a GPU server.

- [Start free](https://dash.allmodels.io/signup)
- [Browse abliterated models](https://allmodels.io/llm/abliterated)
- [Read the AllModels documentation](https://allmodels.io/audio/docs)
- [View the canonical HTML guide](https://allmodels.io/guides/best-abliterated-models)

> AllModels is an independent service and is not affiliated with, endorsed by, or sponsored by the providers whose models it routes to.