Qwen3 30B A3B 2507 Instruct vs Qwen3 VL 30B A3B (Reasoning)
Comparing 2 AI models · 6 benchmarks · Alibaba
Composite Indices
Intelligence, Coding, Math
Standard Benchmarks
Academic and industry benchmarks
Benchmark Winners
Qwen3 30B A3B 2507 Instruct
- MATH 500
Qwen3 VL 30B A3B (Reasoning)
- GPQA
- MMLU Pro
- HLE
- LiveCodeBench
- AIME 2025
| Metric | Al Qwen3 30B A3B 2507 Instruct | Al Qwen3 VL 30B A3B (Reasoning) |
|---|---|---|
| Pricing Per 1M tokens | ||
| Input Cost | $0.20/1M | $0.20/1M |
| Output Cost | $0.80/1M | $2.40/1M |
| Blended Cost 3:1 input/output ratio | $0.35/1M | $0.75/1M |
| Specifications | ||
| Organization Model creator | Alibaba | Alibaba |
| Release Date Launch date | Jul 29, 2025 | Oct 3, 2025 |
| Performance & Speed | ||
| Throughput Output speed | 112.8 tok/s | 104.5 tok/s |
| Time to First Token (TTFT) Initial response delay | 1076ms | 956ms |
| Latency Time to first answer token | 1076ms | 20090ms |
| Composite Indices | ||
| Intelligence Index Overall reasoning capability | 37.0 | 45.3 |
| Coding Index Programming ability | 29.2 | 34.5 |
| Math Index Mathematical reasoning | 66.3 | 82.3 |
| Standard Benchmarks | ||
| GPQA Graduate-level reasoning | 65.9% | 72.0% |
| MMLU Pro Advanced knowledge | 77.7% | 80.7% |
| HLE Hard language evaluation | 6.8% | 8.7% |
| LiveCodeBench Real-world coding tasks | 51.5% | 69.7% |
| MATH 500 Mathematical problems | 97.5% | — |
| AIME 2025 Advanced math competition | 66.3% | 82.3% |
| AIME (Original) Math olympiad problems | 72.7% | — |
| SciCode Scientific code generation | 30.4% | 28.8% |
| LCR Code review capability | 22.7% | 40.7% |
| IFBench Instruction-following | 33.1% | 45.1% |
| TAU-bench v2 Tool use & agentic tasks | 10.2% | 19.9% |
| TerminalBench Hard CLI command generation | 5.7% | 5.0% |
Key Takeaways
Qwen3 30B A3B 2507 Instruct offers the best value at $0.20/1M, making it ideal for high-volume applications and cost-conscious projects.
Qwen3 VL 30B A3B (Reasoning) leads in reasoning capabilities with a 72.0% GPQA score, excelling at complex analytical tasks and problem-solving.
Qwen3 VL 30B A3B (Reasoning) achieves a 34.5 coding index, making it the top choice for software development and code generation tasks.
All models support context windows of ∞+ tokens, suitable for processing lengthy documents and maintaining extended conversations.
When to Choose Each Model
Qwen3 30B A3B 2507 Instruct
- Cost-sensitive applications
- High-volume processing
Qwen3 VL 30B A3B (Reasoning)
- Complex reasoning tasks
- Research & analysis
- Code generation
- Software development
Try Models for Free
Try Qwen3 30B A3B 2507 Instruct for FREE
No credit card or account required.
Try Qwen3 VL 30B A3B (Reasoning) for FREE
No credit card or account required.
Cost Calculator
Qwen3 30B A3B 2507 Instruct Qwen3 VL 30B A3B (Reasoning) Costs are estimates based on API pricing. Actual costs may vary based on caching, batch processing, and volume discounts.
Recommended Comparisons
