NOUMENOpen app
NOUMEN

Qwen · Open weights

Qwen3.5 9B

Added Mar 10, 2026Open weights

Compare models

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

Loading models…

Use in the app

Qwen3.5 9B is a multimodal foundation model from the Qwen 3.5 family, built for efficient reasoning, coding, and visual understanding in a compact 9B architecture. Requests containing video cost $0.17 per million input tokens and $0.25 per million output tokens.

  • Reasoning
  • Tools
  • Vision
  • Video input
Input
Text, Images, Video
Output
Text
Discovery tier
Open weights

Pricing

USD per 1M tokens
Input
$0.05
Output
$0.15

Reference prices as published by the model provider. API example

Specifications

Context window
262.1K
Max output
65.5K
Average output (7 days)
1.4K tokens
Longer than 84% of 465 models

Side by side

Qwen3.5 9B 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…

Qwen3.5 9BQwenThis modelGemini 2.5 Flash 0520 ThinkingGoogleRemove
ProviderQwenGoogle
Discovery tierOpen weightsEfficient
AddedMar 10, 2026May 20, 2025
Limits
Context window262.14K1.05MBest
Max output65.54K65.54K
Average output1.4K tokens—
Pricing, USD per 1M tokens
Input$0.05Best$0.15
Output$0.15Best$3.50
Cache read—$0.015
Performance
Request success (24h)99.8%Best99.6%
Benchmark indexes
Intelligence Index13.3Best13.1
Coding Index23.5—
Math Index—73.3
Capabilities
ReasoningYesYes
ToolsYesNo
VisionYesYes
Video inputYesNo
Reasoning effort—low, medium, high

Public evaluations

Benchmarks

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

  • Qwen3.5 9B
  • Gemini 2.5 Flash 0520 Thinking

Intelligence Index

  • Qwen3.5 9B13.3
  • Gemini 2.5 Flash 0520 Thinking13.1

Qwen3.5 9B: Better than 33% of 326 scored models

3.8Catalog models by score55.1

Indexes

  • Coding Index

    Composite of the coding evaluations

    • Qwen3.5 9B: 23.5
    • Gemini 2.5 Flash 0520 Thinking: —

  • Math Index

    Composite of the math evaluations

    • Qwen3.5 9B: —
    • Gemini 2.5 Flash 0520 Thinking: 73.3

Document reasoning

  • AA-LCRv1.1

    Reasoning across long documents

    • Qwen3.5 9B: 46.0%
    • Gemini 2.5 Flash 0520 Thinking: 65.3%Top

Reasoning and knowledge

  • Humanity’s Last Exam

    Expert-level questions across academic fields

    • Qwen3.5 9B: 9.4%
    • Gemini 2.5 Flash 0520 Thinking: 12.1%Top

  • GPQA Diamond

    Graduate-level science questions

    • Qwen3.5 9B: 78.6%
    • Gemini 2.5 Flash 0520 Thinking: 79.0%Top

  • MMLU-Pro

    Multi-subject knowledge and reasoning

    • Qwen3.5 9B: —
    • Gemini 2.5 Flash 0520 Thinking: 83.2%

  • IFBench

    Following precise output instructions

    • Qwen3.5 9B: 37.8%
    • Gemini 2.5 Flash 0520 Thinking: 50.3%Top

Coding

  • LiveCodeBench

    Recent competitive programming problems

    • Qwen3.5 9B: —
    • Gemini 2.5 Flash 0520 Thinking: 69.5%

Math

  • AIME 2025

    Competition mathematics, 2025 problems

    • Qwen3.5 9B: —
    • Gemini 2.5 Flash 0520 Thinking: 73.3%

  • AIME

    Competition mathematics

    • Qwen3.5 9B: —
    • Gemini 2.5 Flash 0520 Thinking: 82.3%

  • MATH-500

    Competition math problems

    • Qwen3.5 9B: —
    • Gemini 2.5 Flash 0520 Thinking: 98.1%

Legacy evaluations

  • Terminal-Bench Hard

    Earlier terminal task suite

    • Qwen3.5 9B: 18.2%Top
    • Gemini 2.5 Flash 0520 Thinking: 13.6%

  • τ²-Bench Telecom

    Tool use in a customer-support simulation

    • Qwen3.5 9B: 85.1%Top
    • Gemini 2.5 Flash 0520 Thinking: 31.6%

  • AA-LCR (unversioned)

    Long-context reasoning, earlier grading

    • Qwen3.5 9B: 46.0%
    • Gemini 2.5 Flash 0520 Thinking: 65.3%Top

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.

Prompt cache hits
14.3%
Request success (24h)
99.8%
Reliable

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.

Hourly request success · last 72 hoursLowest hour 96.3%
  • Healthy (98% or more)
  • Degraded (90–98%)
  • Down (under 90%)
  • No requests

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": "qwen/qwen3.5-9b",
    "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="qwen/qwen3.5-9b",
    messages=[{"role": "user", "content": "Explain this stack trace."}],
)
print(reply.choices[0].message.content)

Qwen

Other Qwen models, starting with the same family.