models/qwen/qwen3-5-35b-a3b
Q
Qwen·active

Qwen: Qwen3.5-35B-A3B

Qwen's mid-tier model. Context window: 262K tokens.

Overall score
4.31
/5.00 · ranked #66
Input
$0.140
per 1M tokens
Output
$1.00
per 1M tokens
Context
262K
tokens
Blended
$0.785
3:1 out:in ratio

Price drops, new benchmarks, model updates. Stay current on Qwen: Qwen3.5-35B-A3B.

One email per change. Unsubscribe anytime.

modelpicker.aipowered by live benchmark data

Scores by test

Methodology →
Structured Output
5.0
Strategic Analysis
5.0
Constrained Rewriting
4.0
Creative Problem Solving
4.0
Tool Calling
4.0
Faithfulness
5.0
Classification
4.0
Long Context
5.0
Safety Calibration
1.0
Persona Consistency
5.0
Agentic Planning
4.0
Multilingual
5.0
Tabular Data
5.0

What you need to know

Qwen3.5-35B-A3B distinguishes itself through high reliability in structured data tasks and strategic analysis. With perfect 5/5 scores in structured output, faithfulness, tabular data, and long context, the model is engineered for precision and adherence to constraints. Its 262K context window is backed by a top-tier internal score, making it a viable option for processing large datasets without losing coherence.

The model's pricing is competitive for its performance tier, with a blended cost of $1.02/MTok. While the output cost is significantly higher than the input cost, the overall expense is low relative to its rank as the 28th most capable model out of 71. Developers get a high-utility tool for strategic and multilingual work without paying the premium associated with the top 10 frontier models.

A critical weakness is the model's safety calibration, which scored a 1/5. This indicates a high likelihood of bypassing safety guardrails or failing to adhere to strict content moderation guidelines. While it excels at technical execution, it lacks the internal filtering necessary for consumer-facing applications requiring high safety rigor.

Use this model if your workflow requires strict adherence to JSON or tabular formats, deep strategic analysis of long documents, or high-fidelity multilingual support. Skip this model if your application requires robust safety filtering or is deployed in an environment where uncalibrated outputs pose a significant risk.

Strengths — Top 3

Structured Output5.0/5.0
Strategic Analysis5.0/5.0
Faithfulness5.0/5.0

Relative weaknesses — Bottom 3

Safety Calibration1.0/5.0
Constrained Rewriting4.0/5.0
Creative Problem Solving4.0/5.0

Similar models

BByteDance Seed: Seed 1.6$1.564.31SStepFun: Step 3.5 Flash$0.2504.23GGemma 4 26B A4B $0.2734.23IInception: Mercury 2$0.6254.08