Qwen3 30B A3B
Qwen3, the latest generation in the Qwen large language model series, features both dense and mixture-of-experts (MoE) architectures to excel in reasoning, multilingual support, and advanced agent tasks. Its unique ability to switch seamlessly between a thinking mode for complex reasoning and a non-thinking mode for efficient dialogue ensures versatile, high-quality performance. Significantly outperforming prior models like QwQ and Qwen2.5, Qwen3 delivers superior mathematics, coding, commonsense reasoning, creative writing, and interactive dialogue capabilities. The Qwen3-30B-A3B variant includes 30.5 billion parameters (3.3 billion activated), 48 layers, 128 experts (8 activated per task), and supports up to 131K token contexts with YaRN, setting a new standard among open-source models.
- Input / output price
- $0.12 / $0.50 per 1M
- Context
- 41K tokens
- Released
- Apr 29, 2025
Paste into Claude Code, Cursor, or Codex
Help me set up the Respan gateway: fetch and follow https:/
Providers
3 providers, cheapest first, OpenRouter last, priced per 1M tokens after gateway discounts.
| Provider | Input | Output |
|---|---|---|
DeepInfra deepinfra/qwen3-30b-a3b | $0.12/M | $0.50/M |
Inference.net inference_net/qwen3-30b-a3b | $0.12/M | $0.50/M |
OpenRouter openrouter/qwen3-30b-a3b | $0.12/M | $0.50/M |
Is a provider down? OpenRouter status
Pricing
Averaged across 3 providers, in USD per 1M tokens after gateway discounts.
- Average input
- $0.12 per 1M
- Average output
- $0.50 per 1M
Cost by volume
Total cost from 0 to 1T tokens.
Throughput and latency
30-day averages from requests through the Respan gateway. Under each figure, every model on the gateway by its average, this one in blue.
Output speed
Higher is better16tokens/sacross providers
Response latency
Lower is better0.7sacross providers
Benchmarks
The Respan Index plus 11 published scores from 5 sources, each linked to where it was published.
14.9/ 100
Respan Index#161 of 176
- 161Qwen3 30B-A3B14.9
- 160QwQ-32B15.0
- 161Granite 4.2 8B14.9
Published scores
| Benchmark | Score | Source | Published |
|---|---|---|---|
| LiveCodeBench v5arxiv.org | 62.6% | arxiv.org | May 14, 2025 |
| WeirdMLHåvard Tveit Ihle | 29.8% | Håvard Tveit Ihle | Oct 31, 2026 |
| Benchmark | Score | Source | Published |
|---|---|---|---|
| GPQA Diamondarxiv.org | 65.8% | arxiv.org | May 14, 2025 |
| GPQA DiamondEpoch AI | 61.7% | Epoch AI | Oct 31, 2026 |
| Benchmark | Score | Source | Published |
|---|---|---|---|
| AIME 2024arxiv.org | 80.4% | arxiv.org | May 14, 2025 |
| AIME 2025arxiv.org | 70.9% | arxiv.org | May 14, 2025 |
| AIME 2025MathArena | 70% | MathArena | Oct 31, 2026 |
| OTIS Mock AIME 2024-2025Epoch AI | 62.8% | Epoch AI | Oct 31, 2026 |
| Benchmark | Score | Source | Published |
|---|---|---|---|
| LMArena EloLMArena | 1327 Elo · #163 of 191 | LMArena | Nov 19, 2026 |
How Qwen compares: DeepSeek vs Qwen, Claude vs Qwen, Gemini vs Qwen
Compare Qwen3 30B A3B: Qwen3 30B-A3B vs Qwen3.8-Max, Qwen3 30B-A3B vs Qwen3.8-Flash-Next, Qwen3 30B-A3B vs Qwen3.7-Max, Qwen3 30B-A3B vs Qwen3.8 27B, Qwen3 30B-A3B vs Qwen3.6-Max-Preview, Qwen3 30B-A3B vs Qwen3.7-Plus
Apps
Use Qwen3 30B A3B in the tools you already run. Each guide points the app at the Respan gateway.
Setup guides for more apps are in the integration docs.
Qwen3 30B A3B FAQ
Short answers, built from the data on this page.
Qwen3 30B A3B is a model from Qwen, released on Apr 29, 2025. Qwen3, the latest generation in the Qwen large language model series, features both dense and mixture-of-experts (MoE) architectures to excel in reasoning, multilingual support, and advanced agent tasks. Its unique ability to switch seamlessly between a thinking mode for complex reasoning and a non-thinking mode for efficient dialogue ensures versatile, high-quality performance. Significantly outperforming prior models like QwQ and Qwen2.5, Qwen3 delivers superior mathematics, coding, commonsense reasoning, creative writing, and interactive dialogue capabilities. The Qwen3-30B-A3B variant includes 30.5 billion parameters (3.3 billion activated), 48 layers, 128 experts (8 activated per task), and supports up to 131K token contexts with YaRN, setting a new standard among open-source models.
On the Respan gateway, Qwen3 30B A3B costs $0.12 per 1M input tokens and $0.50 per 1M output tokens on every provider, after gateway discounts.
Qwen3 30B A3B is available through 3 providers: DeepInfra, Inference.net, and OpenRouter.
Qwen3 30B A3B accepts up to 41K tokens of context on its largest provider route.
Point any OpenAI-compatible SDK at https://gateway.respan.ai/v1 with your Respan API key and set the model to "qwen3-30b-a3b". To set it up with a coding agent, copy the agent setup prompt at the top of this page.
Qwen3 30B A3B scores 14.9 out of 100 on the Respan Index v1.1, ranking #161 of 176 models. By category: Coding 2.7, Agents 11.1 (estimated), Reasoning 20.2, Knowledge & language 10.0 (estimated), and Human preference 35.8. The Respan Index combines independent and lab-reported benchmark results into one score, using benchmark data collected by Respan.
Qwen3 30B A3B has 11 published benchmark scores, including LiveCodeBench v5 62.6%, BFCL-v3 69.1%, and GPQA Diamond 65.8%. Each score links to where it was published; the data is collected by Respan.
Certified where it counts.
Built to meet the security and privacy standards that enterprise and healthcare teams require.
ISO 27001
The internationally recognized standard for information security management.
SOC 2
Secure, compliant management of your data across all of our systems.
GDPR
Operated under GDPR, the world's strictest standard for data privacy.
HIPAA
HIPAA compliant, with a BAA available for healthcare teams.