NOUMENOpen app
NOUMEN

Qwen · Open weights

Qwen 2.5 32B Abliterated

Added Jan 6, 2025Open weights

Compare models

Choose up to 3 models to compare.1 of 3 chosen

Loading models…

Use in the app

Uncensored version of Qwen 2.5 32B Instruct with restrictions removed.

Input
Text
Output
Text
Discovery tier
Open weights

Pricing

USD per 1M tokens
Input
$0.70
Output
$0.70

Reference prices as published by the model provider. API example

Specifications

Context window
32.8K
Max output
8.2K
Average output (7 days)
806 tokens
Longer than 65% of 465 models

Side by side

Qwen 2.5 32B Abliterated compared

Specifications, prices, speed and benchmark scores next to each other. Benchmarks below show every compared model too.

Compare models

Choose up to 3 models to compare.1 of 3 chosen

Loading models…

Qwen 2.5 32B AbliteratedQwenThis modelGLM 5.1 TEEZ.aiRemove
ProviderQwenZ.ai
Discovery tierOpen weightsOpen weights
AddedJan 6, 2025Apr 20, 2026
Limits
Context window32.77K202.75KBest
Max output8.19K65.54KBest
Average output806 tokens675 tokens
Pricing, USD per 1M tokens
Input$0.70$0.70
Output$0.70Best$2.50
Cache read—$0.18
Performance
Throughput19.6 tok/s—
Time to first token2.1 s—
Request success (24h)60.4%100.0%Best
Benchmark indexes
Intelligence Index—26.1
Coding Index—55.8
Capabilities
ReasoningNoYes
ToolsNoYes
Reasoning effort—none, high

Public evaluations

Benchmarks

Independent evaluation results, with how each score ranks among the models in the Noumen catalog.

  • Qwen 2.5 32B Abliterated
  • GLM 5.1 TEE

Intelligence Index

  • Qwen 2.5 32B Abliterated—
  • GLM 5.1 TEE26.1
3.8Catalog models by score55.1

Indexes

  • Coding Index

    Composite of the coding evaluations

    • Qwen 2.5 32B Abliterated: —
    • GLM 5.1 TEE: 55.8

Agentic work

  • AutomationBench-AA

    Workflow automation, penalized for guardrail violations

    • Qwen 2.5 32B Abliterated: —
    • GLM 5.1 TEE: 20.3%

  • AutomationBench-AA tasks completed

    Workflows finished end to end without a guardrail violation

    • Qwen 2.5 32B Abliterated: —
    • GLM 5.1 TEE: 3.0%

  • AA-Briefcasev1.1

    Agentic knowledge work, rated head to head

    • Qwen 2.5 32B Abliterated: —
    • GLM 5.1 TEE: 964 Elo

  • GDPval-AAv2.1

    Economically valuable occupational tasks, rated head to head

    • Qwen 2.5 32B Abliterated: —
    • GLM 5.1 TEE: 1,117 Elo

Document reasoning

  • GDP.pdf

    Professional reasoning over PDF documents, all-pass rate

    • Qwen 2.5 32B Abliterated: —
    • GLM 5.1 TEE: 8.4%

  • AA-LCRv1.1

    Reasoning across long documents

    • Qwen 2.5 32B Abliterated: —
    • GLM 5.1 TEE: 73.7%

Reasoning and knowledge

  • Humanity’s Last Exam

    Expert-level questions across academic fields

    • Qwen 2.5 32B Abliterated: —
    • GLM 5.1 TEE: 30.1%

  • GPQA Diamond

    Graduate-level science questions

    • Qwen 2.5 32B Abliterated: —
    • GLM 5.1 TEE: 86.8%

  • IFBench

    Following precise output instructions

    • Qwen 2.5 32B Abliterated: —
    • GLM 5.1 TEE: 76.3%

Coding

  • Terminal-Benchv4.0

    Practical software engineering in a terminal

    • Qwen 2.5 32B Abliterated: —
    • GLM 5.1 TEE: 2.0%

  • SciCode

    Scientific computing in Python

    • Qwen 2.5 32B Abliterated: —
    • GLM 5.1 TEE: 44.8%

Legacy evaluations

  • Terminal-Bench Hard

    Earlier terminal task suite

    • Qwen 2.5 32B Abliterated: —
    • GLM 5.1 TEE: 43.2%

  • τ²-Bench Telecom

    Tool use in a customer-support simulation

    • Qwen 2.5 32B Abliterated: —
    • GLM 5.1 TEE: 97.7%

  • AA-LCR (unversioned)

    Long-context reasoning, earlier grading

    • Qwen 2.5 32B Abliterated: —
    • GLM 5.1 TEE: 73.7%

Scores from Artificial Analysis, updated Oct 26, 2026. Artificial Analysis

Live routing data

Performance and reliability

Recent averages across requests routed for this model. Your own speed depends on prompt length, reasoning effort and load.

Throughput
19.6 tok/s
Faster than 12% of 251 models
Time to first token
2.1 s
Quicker than 72% of 251 models
Prompt cache hits
0.0%
Request success (24h)
60.4%

Cache hits are the share of input tokens recently served from a prompt cache across all requests. They are not a forecast of your own hit rate.

Developers

Use it through the API

Noumen’s API is OpenAI-compatible: any Chat Completions client works with a base URL and a key.

Read the quickstart
curl https://www.noumen.xyz/api/v1/chat/completions \
  -H "Authorization: Bearer $NOUMEN_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "huihui-ai/Qwen2.5-32B-Instruct-abliterated",
    "messages": [{ "role": "user", "content": "Explain this stack trace." }]
  }'
import os
from openai import OpenAI

client = OpenAI(base_url="https://www.noumen.xyz/api/v1", api_key=os.environ["NOUMEN_API_KEY"])
reply = client.chat.completions.create(
    model="huihui-ai/Qwen2.5-32B-Instruct-abliterated",
    messages=[{"role": "user", "content": "Explain this stack trace."}],
)
print(reply.choices[0].message.content)

Qwen

Other Qwen models, starting with the same family.