NOUMENOpen app
NOUMEN

Qwen · Open weights

Qwen: QvQ Max

Added Mar 28, 2025Open weights

Compare models

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

Loading models…

Use in the app

QvQ Max is the top model of the Qwen series. QvQ Max is capable of thinking and reasoning, can achieve significantly enhanced performance especially on hard problems.

  • Vision
Input
Text, Images
Output
Text
Discovery tier
Open weights

Pricing

USD per 1M tokens
Input
$1.20
Output
$4.80

Reference prices as published by the model provider. API example

Specifications

Context window
128K
Max output
8.2K

Side by side

Qwen: QvQ Max 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: QvQ MaxQwenThis modelKimi K3MoonshotRemove
ProviderQwenMoonshot
Discovery tierOpen weightsOpen weights
AddedMar 28, 2025Jul 16, 2026
Limits
Context window128K1.05MBest
Max output8.19K—
Average output—1.1K tokens
Pricing, USD per 1M tokens
Input$1.20$1.20
Output$4.80Best$5.50
Cache read—$0.20
Performance
Throughput—137.3 tok/s
Time to first token—10.6 s
Request success (24h)—99.9%
Benchmark indexes
Intelligence Index—43.6
Coding Index—76.2
Capabilities
ReasoningNoYes
ToolsNoYes
VisionYesYes
Structured outputNoYes
Video inputNoYes
Reasoning effort—low, high, max

Public evaluations

Benchmarks

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

  • Qwen: QvQ Max
  • Kimi K3

Intelligence Index

  • Qwen: QvQ Max—
  • Kimi K343.6
3.8Catalog models by score55.1

Indexes

  • Coding Index

    Composite of the coding evaluations

    • Qwen: QvQ Max: —
    • Kimi K3: 76.2

Agentic work

  • AutomationBench-AA

    Workflow automation, penalized for guardrail violations

    • Qwen: QvQ Max: —
    • Kimi K3: 58.3%

  • AutomationBench-AA tasks completed

    Workflows finished end to end without a guardrail violation

    • Qwen: QvQ Max: —
    • Kimi K3: 24.8%

  • Harvey LAB-AAv1.1

    Legal tasks passed in full, with no material hallucination

    • Qwen: QvQ Max: —
    • Kimi K3: 5.3%

  • AA-Briefcasev1.1

    Agentic knowledge work, rated head to head

    • Qwen: QvQ Max: —
    • Kimi K3: 1,499 Elo

  • GDPval-AAv2.1

    Economically valuable occupational tasks, rated head to head

    • Qwen: QvQ Max: —
    • Kimi K3: 1,534 Elo

Document reasoning

  • GDP.pdf

    Professional reasoning over PDF documents, all-pass rate

    • Qwen: QvQ Max: —
    • Kimi K3: 22.0%

  • AA-LCRv1.1

    Reasoning across long documents

    • Qwen: QvQ Max: —
    • Kimi K3: 88.7%

  • MLCR-AA

    Long-context reasoning over medical records

    • Qwen: QvQ Max: —
    • Kimi K3: 21.7%

Reasoning and knowledge

  • Humanity’s Last Exam

    Expert-level questions across academic fields

    • Qwen: QvQ Max: —
    • Kimi K3: 46.9%

  • GPQA Diamond

    Graduate-level science questions

    • Qwen: QvQ Max: —
    • Kimi K3: 93.5%

Coding

  • Terminal-Benchv4.0

    Practical software engineering in a terminal

    • Qwen: QvQ Max: —
    • Kimi K3: 12.6%

  • SciCode

    Scientific computing in Python

    • Qwen: QvQ Max: —
    • Kimi K3: 59.5%

Legacy evaluations

  • AA-LCR (unversioned)

    Long-context reasoning, earlier grading

    • Qwen: QvQ Max: —
    • Kimi K3: 88.7%

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

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": "qvq-max",
    "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="qvq-max",
    messages=[{"role": "user", "content": "Explain this stack trace."}],
)
print(reply.choices[0].message.content)

Qwen

Other Qwen models, starting with the same family.