Qwen3 30B A3B (Reasoning)
Intelligence
15.3
#330/523
Coding
11
#318/429
Math
72.3
#91/265
Speed
69 tok/s
TTFT: 1.10s
Pricing
$0.20 / $2.40
per 1M tokens (in/out)
Qwen3-30B-A3B is a smaller Mixture-of-Experts (MoE) model from the Qwen3 series by Alibaba, with 30.5 billion total parameters and 3.3 billion activated parameters. Features hybrid thinking/non-thinking modes, support for 119 languages, and enhanced agent capabilities. It aims to outperform previous models like QwQ-32B while using significantly fewer activated parameters.
Qwen3 30B A3B (Reasoning) is Alibaba’s model designed for advanced reasoning tasks. It processes at 68.893 tokens per second and is priced at $0.2 per million input tokens, targeting professional users.
When to Use Qwen3 30B A3B (Reasoning)
✓ Best For
- Complex reasoning tasks.
- Mathematical problem solving.
- Coding assistance.
✗ Not Ideal For
- Basic conversational applications.
- High-speed real-time processing.
How Qwen3 30B A3B (Reasoning) Compares
Intelligence Index · Higher is better
StepFunPerplexityAlibabaGoogleMistral
Benchmark Profile
Coding Index
MistralZ AIAlibabaAmazon
Output Speed · tok/s
AlibabaZ AIMiniMax
Math Index
Z AIAlibabaInclusionAI
Intelligence · Coding · Math
Intelligence
Coding
Math
All Benchmark Scores (14)
| Benchmark | Score | |
|---|---|---|
| Intelligence Index | 15.3 | |
| Coding Index | 11 | |
| Math Index | 72.3 | |
| MMLU-Pro | 777% | |
| GPQA | 616% | |
| LiveCodeBench | 506% | |
| HLE | 66% | |
| SciCode | 28.5% | |
| IFBench | 41.5% | |
| TerminalBench Hard | 2.3% | |
| Tau2 | 26% | |
| AIME | 75.3% | |
| AIME 2025 | 72.3% | |
| MATH 500 | 95.9% |
Data: Artificial Analysis · Updated: April 10, 2026
Frequently Asked Questions (15)
When was Qwen3 30B A3B (Reasoning) released?
Qwen3 30B A3B (Reasoning) was released on April 28, 2025.
Who created Qwen3 30B A3B (Reasoning)?
Qwen3 30B A3B (Reasoning) was created by Alibaba.
How intelligent is Qwen3 30B A3B (Reasoning)?
Qwen3 30B A3B (Reasoning) scores 15 on the Artificial Analysis Intelligence Index, placing it above average among other open weight models of similar size (median: 15).
How fast is Qwen3 30B A3B (Reasoning)?
Qwen3 30B A3B (Reasoning) generates output at 65.3 tokens per second (based on Alibaba's API), which is below average compared to other open weight models of similar size (median: 97.2 t/s).
What is the latency of Qwen3 30B A3B (Reasoning)?
Qwen3 30B A3B (Reasoning) has a time to first token (TTFT) of 2.42s (based on Alibaba's API), which is at the higher end compared to other open weight models of similar size (median: 1.90s).
How much does Qwen3 30B A3B (Reasoning) cost?
Qwen3 30B A3B (Reasoning) costs $0.20 per 1M input tokens (somewhat higher than average, median: $0.18) and $2.40 per 1M output tokens (at the higher end, median: $0.40), based on Alibaba's API.
What is Qwen3 30B A3B (Reasoning) API pricing?
Qwen3 30B A3B (Reasoning) costs $0.20 per 1M input tokens and $2.40 per 1M output tokens (based on Alibaba's API). For a blended rate (3:1 input to output ratio), this is $0.75 per 1M tokens. Pricing may vary by provider.
How verbose is Qwen3 30B A3B (Reasoning)?
When evaluated on the Intelligence Index, Qwen3 30B A3B (Reasoning) generated 32M output tokens, which is somewhat higher than average compared to other open weight models of similar size (median: 23M).
Is Qwen3 30B A3B (Reasoning) a reasoning model?
Yes, Qwen3 30B A3B (Reasoning) is a reasoning model. It uses extended thinking or chain-of-thought reasoning to work through complex problems before providing an answer.
What input modalities does Qwen3 30B A3B (Reasoning) support?
Qwen3 30B A3B (Reasoning) supports text input.
What output modalities does Qwen3 30B A3B (Reasoning) support?
Qwen3 30B A3B (Reasoning) supports text output.
Can Qwen3 30B A3B (Reasoning) process images?
No, Qwen3 30B A3B (Reasoning) does not support image input. It can only process text.
Is Qwen3 30B A3B (Reasoning) multimodal?
No, Qwen3 30B A3B (Reasoning) is not multimodal. It only supports text input.
What is the context window of Qwen3 30B A3B (Reasoning)?
Qwen3 30B A3B (Reasoning) has a context window of 33k tokens. This determines how much text and conversation history the model can process in a single request.
Is Qwen3 30B A3B (Reasoning) open source?
Yes, Qwen3 30B A3B (Reasoning) is open weights. The model weights are publicly available and can be downloaded for self-hosting.
