Intelligence
21
score
Qwen3 VL 235B A22B (Reasoning) scores 21 (estimated) on the Artificial Analysis Intelligence Index, …
vs class avg: 25
Alibaba•Open weights•September 23, 2025
Qwen3 VL 235B A22B
Intelligence, Performance & Price Analysis
Source:Artificial Analysis scrapeType:Open weights
Speed
54.5
output tokens/sec
Qwen3 VL 235B A22B (Reasoning) generates output at 54.5 tokens per second (based on Alibaba's API), …
Latency
3.02s
TTFT
Qwen3 VL 235B A22B (Reasoning) has a time to first token (TTFT) of 3. …
Input Price#68/93
$0.70
/ 1M tokens
Qwen3 VL 235B A22B (Reasoning) costs $0.70 per 1M input tokens (somewhat higher than average, …
Output Price#91/93
$8.40
/ 1M tokens
Qwen3 VL 235B A22B (Reasoning) costs $0.70 per 1M input tokens (somewhat higher than average, …
Decision Overview
Ethen's intelligence-driven routing assessment for this model
Overall verdict
Strong general-purpose model
Suitable for most production tasks, but high-volume or repetitive work should still be compared against cheaper routes.
Best for
- Code generation
- Content analysis
- Structured extraction
- Customer-facing chat
- High-value tasks where quality outweighs cost
Weak or unsuitable
- Extremely long-context tasks
- Ultra-low-latency requirements
- Routine high-volume generation
- Simple extraction or labeling
Routing recommendation
1
Lower-cost models for repetitive or high-volume work
2
Cache-backed reuse for repeated prompts
Cost pressure
Premium pricing — $0.70 / M input, $8.40 / M output. Use cheaper routes for repetitive or low-risk work.
Model-fit matrix
Use this model where capability pays for itself; route away where volume, latency, or budget dominate.
| Use case | Fit | Reason | Routing note |
|---|---|---|---|
| Complex reasoning & agentic workflows | Strong | Intelligence score 21 supports capable reasoning, but very hard tasks may benefit from higher-tier models. | Good for most complex tasks; consider a frontier model for the hardest 10%. |
| High-volume chat & customer-facing | Strong | Output speed 54.5 tokens/sec is adequate for chat. | Suitable for chat; monitor latency under concurrent load. |
| Latency-sensitive applications | Moderate | TTFT 3.02s — latency may be noticeable in interactive use. | Consider routing latency-critical paths to faster models. |
| Cost-sensitive pipelines | Moderate | Output pricing at $8.40 is premium. Route high-volume simple tasks to cheaper alternatives. | Use only for high-value tasks; route simple queries to budget models. |
Cost pressure
HighPricing is premium — input $0.70, output $8.40. This model is expensive for high-volume or output-heavy workloads.
Cheaper substitutes
- Lower-cost models for repetitive or high-volume work
- Cache-backed reuse for repeated prompts
Route away when
- Output-heavy tasks
- High-frequency chat at scale
- Simple queries that cheaper models handle well
Highlights
Key benchmark results at a glance.
Intelligence
Intelligence position
Artificial Analysis Intelligence Index · Higher is better · Evaluation results measured independently by Artificial Analysis
Qwen3 VL 235B A22B (Reasoning) scores 21 (estimated) on the Artificial Analysis Intelligence Index, placing it below average among other open weight models of similar size (median: 25).
Speed
Speed position
Output tokens per second · Higher is better · Evaluation results measured independently by Artificial Analysis
Qwen3 VL 235B A22B (Reasoning) generates output at 54.5 tokens per second (based on Alibaba's API), which is below average compared to other open weight models of similar size (median: 61.7 t/s).
Context, Cost & Pricing
Price/value position
Weighted average cost (USD) per Intelligence Index task · Lower is better · Evaluation results measured independently by Artificial Analysis
Pricing is premium — input $0.70, output $8.40. This model is expensive for high-volume or output-heavy workloads.
Intelligence & Quality
Benchmark scores and quality indices measuring model capability.
AA-Omniscience Index
AA-Omniscience Index (higher is better) measures knowledge reliability and hallucination. It rewards correct answers, penalizes hallucinations, and has no penalty for refusing to answer. Scores range from -100 to 100, where 0 means as many correct as incorrect answers, and negative scores mean more incorrect than correct. · Evaluation results measured independently by Artificial Analysis
Source and methodology are summarized below
Chart source and provenance are listed in Methodology & sources below.
Bars intentionally show many models at once so the page feels like a benchmark report rather than a dashboard.
Artificial Analysis Intelligence Index by Open Weights / Proprietary
Artificial Analysis Intelligence Index v4.1 incorporates 9 evaluations: GDPval-AA v2, 𝜏³-Banking, Terminal-Bench v2.1, SciCode, Humanity's Last Exam, GPQA Diamond, CritPt, AA-Omniscience, AA-LCR · Evaluation results measured independently by Artificial Analysis
Source and methodology are summarized below
Chart source and provenance are listed in Methodology & sources below.
Higher values indicate stronger overall capability. Bars show how models compare on composite intelligence metrics.
Artificial Analysis Intelligence Index
Artificial Analysis Intelligence Index v4.1 incorporates 9 evaluations: GDPval-AA v2, 𝜏³-Banking, Terminal-Bench v2.1, SciCode, Humanity's Last Exam, GPQA Diamond, CritPt, AA-Omniscience, AA-LCR · Evaluation results measured independently by Artificial Analysis
Source and methodology are summarized below
Chart source and provenance are listed in Methodology & sources below.
Higher values indicate stronger overall capability. Bars show how models compare on composite intelligence metrics.
Artificial Analysis Openness Index: Score
Openness Index assesses model openness on a 0 to 100 normalized scale (higher is more open) · Evaluation results measured independently by Artificial Analysis
Source and methodology are summarized below
Chart source and provenance are listed in Methodology & sources below.
Bars intentionally show many models at once so the page feels like a benchmark report rather than a dashboard.
Intelligence
Artificial Analysis Intelligence Index · Higher is better · Evaluation results measured independently by Artificial Analysis
Source and methodology are summarized below
Chart source and provenance are listed in Methodology & sources below.
Higher values indicate stronger overall capability. Bars show how models compare on composite intelligence metrics.
Speed & Throughput
Output tokens per second and generation throughput.
Output Speed
Output tokens per second · Higher is better · Evaluation results measured independently by Artificial Analysis
Source and methodology are summarized below
Chart source and provenance are listed in Methodology & sources below.
Speed should be read together with latency and cost; fast output does not always mean better user-perceived performance.
Speed
Output tokens per second · Higher is better · Evaluation results measured independently by Artificial Analysis
Source and methodology are summarized below
Chart source and provenance are listed in Methodology & sources below.
Speed should be read together with latency and cost; fast output does not always mean better user-perceived performance.
Latency & Response Time
Time to first token and end-to-end response latency.
End-to-End Response Time
Seconds to output 500 tokens, including reasoning model 'thinking' time · Lower is better · Evaluation results measured independently by Artificial Analysis
Source and methodology are summarized below
Chart source and provenance are listed in Methodology & sources below.
Bars intentionally show many models at once so the page feels like a benchmark report rather than a dashboard.
Latency: Time To First Answer Token
Seconds to first answer token received · Accounts for reasoning model 'thinking' time · Evaluation results measured independently by Artificial Analysis
Source and methodology are summarized below
Chart source and provenance are listed in Methodology & sources below.
Latency includes thinking / reasoning time when applicable. Lower is better for interactive use cases.
Context, Cost & Pricing
Context window size, cost per task, and token pricing.
Context Window
Context window: tokens limit · Higher is better · Evaluation results measured independently by Artificial Analysis
Source and methodology are summarized below
Chart source and provenance are listed in Methodology & sources below.
Larger context windows enable complex multi-turn tasks but may increase latency and cost for long inputs.
Cost per Intelligence Index Task
Weighted average cost (USD) per Artificial Analysis Intelligence Index task, segmented by token type. Lower is better · Evaluation results measured independently by Artificial Analysis
Source and methodology are summarized below
Chart source and provenance are listed in Methodology & sources below.
Higher values indicate stronger overall capability. Bars show how models compare on composite intelligence metrics.
Ethen Gateway should route away from this model when the task is simple, repetitive, or likely to generate many output tokens.
Cost per Task
Weighted average cost (USD) per Intelligence Index task · Lower is better · Evaluation results measured independently by Artificial Analysis
Source and methodology are summarized below
Chart source and provenance are listed in Methodology & sources below.
Ethen Gateway should route away from this model when the task is simple, repetitive, or likely to generate many output tokens.
Cost to Run Artificial Analysis Intelligence Index
Cost (USD) to run all evaluations in the Artificial Analysis Intelligence Index · Evaluation results measured independently by Artificial Analysis
Source and methodology are summarized below
Chart source and provenance are listed in Methodology & sources below.
Higher values indicate stronger overall capability. Bars show how models compare on composite intelligence metrics.
Ethen Gateway should route away from this model when the task is simple, repetitive, or likely to generate many output tokens.
Pricing: Cache Hit, Input, and Output
Price (USD per M Tokens) · Evaluation results measured independently by Artificial Analysis
Source and methodology are summarized below
Chart source and provenance are listed in Methodology & sources below.
Bars intentionally show many models at once so the page feels like a benchmark report rather than a dashboard.
Time per Intelligence Index Task
Weighted average decode time (minutes) per task; excludes TTFT and overhead time · Lower is better · Evaluation results measured independently by Artificial Analysis
Source and methodology are summarized below
Chart source and provenance are listed in Methodology & sources below.
Higher values indicate stronger overall capability. Bars show how models compare on composite intelligence metrics.
Architecture & Scale
Model size, parameters, and architectural details.
Model Size: Total and Active Parameters
Comparison between total model parameters and parameters active during inference · Evaluation results measured independently by Artificial Analysis
Source and methodology are summarized below
Chart source and provenance are listed in Methodology & sources below.
Larger models generally perform better but cost more to serve. Active parameter count matters for inference efficiency.
Specifications
Technical Specifications
Qwen3 VL 235B A22B (Reasoning) scores 21 (estimated) on the Artificial Analysis Intelligence Index, placing it below average among other open weight models of similar size (median: 25). Qwen3 VL 235B A22B (Reasoning) generates output at 54.5 tokens per second (based on Alibaba's API), which is below average compared to other open weight models of similar size (median: 61.7 t/s). Qwen3 VL 235B A22B (Reasoning) costs $0.70 per 1M input tokens (somewhat higher than average, median: $0.59) and $8.40 per 1M output tokens (at the higher end, median: $2.20), based on Alibaba's API.
Qwen3 VL 235B A22B
- Model type
- Open weights
- Reasoning
- Yes
- Input modalities
- Qwen3 VL 235B A22B (Reasoning) supports text and image input.
- Output modalities
- Qwen3 VL 235B A22B (Reasoning) supports text output.
- Context window
- 260k tokens
- Open weights / source
- Yes, Qwen3 VL 235B A22B (Reasoning) is open weights. The model weights are publicly available and can be downloaded for self-hosting.
- Parameters
- Qwen3 VL 235B A22B (Reasoning) has 235 billion parameters (22 billion active).
- Active parameters
- Qwen3 VL 235B A22B (Reasoning) is a Mixture of Experts (MoE) model with 235 billion total parameters, but only 22 billion active parameters are used during inference.
- License
- Qwen3 VL 235B A22B (Reasoning) is released under the Apache 2.0 license. This license allows commercial use.
- API availability
- Yes, Qwen3 VL 235B A22B (Reasoning) is available via API through 2 providers.
Methodology & Provenance
This page is rendered from the normalized profile and page JSON for Qwen3 VL 235B A22B.
Benchmark values are preserved as normalized; only layout, disclosure ordering, and typography are adjusted for readability.
Source file: md/scrape/artificial_analysis_model_pages/artificialanalysis_ai__models__qwen3-vl-235b-a22b-reasoning__1e7def769f.html · Canonical URL: https://artificialanalysis.ai/models/qwen3-vl-235b-a22b-reasoning
Normalized at: 2026-07-08T03:24:56+00:00 Technical specs, comparison summary, and FAQs are passed through from the normalized data without rewriting.
Ranks shown on summary cards come directly from the normalized profile. No rank or class display is synthesized in the renderer.
Visible FAQ answers and FAQ JSON-LD are emitted from the same normalized FAQ array, so structured data stays aligned with the rendered accordion.
Time per Intelligence Index Task: https://artificialanalysis.ai/models/qwen3-vl-235b-a22b-reasoning
Frequently Asked Questions
Qwen3 VL 235B A22B (Reasoning) was released on September 23, 2025.
Qwen3 VL 235B A22B (Reasoning) was created by Alibaba.
Qwen3 VL 235B A22B (Reasoning) scores 21 (estimated) on the Artificial Analysis Intelligence Index, placing it below average among other open weight models of similar size (median: 25).