Qwen3 235B A22B 2507 (Reasoning) vs Doubao Seed Code
Comparing 2 AI models · 6 benchmarks · Alibaba, ByteDance Seed
Composite Indices
Intelligence, Coding, Math
Standard Benchmarks
Academic and industry benchmarks
Benchmark Winners
Qwen3 235B A22B 2507 (Reasoning)
- GPQA
- HLE
- LiveCodeBench
- MATH 500
- AIME 2025
Doubao Seed Code
- MMLU Pro
| Metric | Al Qwen3 235B A22B 2507 (Reasoning) | By Doubao Seed Code |
|---|---|---|
| Pricing Per 1M tokens | ||
| Input Cost | $0.70/1M | $0.17/1M |
| Output Cost | $8.40/1M | $1.12/1M |
| Blended Cost 3:1 input/output ratio | $2.63/1M | $0.41/1M |
| Specifications | ||
| Organization Model creator | Alibaba | ByteDance Seed |
| Release Date Launch date | Jul 25, 2025 | Nov 11, 2025 |
| Performance & Speed | ||
| Throughput Output speed | 82.6 tok/s | — |
| Time to First Token (TTFT) Initial response delay | 1136ms | — |
| Latency Time to first answer token | 25354ms | — |
| Composite Indices | ||
| Intelligence Index Overall reasoning capability | 57.5 | 57.1 |
| Coding Index Programming ability | 44.6 | 47.4 |
| Math Index Mathematical reasoning | 91.0 | 79.3 |
| Standard Benchmarks | ||
| GPQA Graduate-level reasoning | 79.0% | 76.4% |
| MMLU Pro Advanced knowledge | 84.3% | 85.4% |
| HLE Hard language evaluation | 15.0% | 13.3% |
| LiveCodeBench Real-world coding tasks | 78.8% | 76.6% |
| MATH 500 Mathematical problems | 98.4% | — |
| AIME 2025 Advanced math competition | 91.0% | 79.3% |
| AIME (Original) Math olympiad problems | 94.0% | — |
| SciCode Scientific code generation | 42.4% | 40.7% |
| LCR Code review capability | 67.0% | 65.3% |
| IFBench Instruction-following | 51.2% | 51.4% |
| TAU-bench v2 Tool use & agentic tasks | 53.2% | 58.2% |
| TerminalBench Hard CLI command generation | 12.8% | 24.8% |
Key Takeaways
Doubao Seed Code offers the best value at $0.17/1M, making it ideal for high-volume applications and cost-conscious projects.
Qwen3 235B A22B 2507 (Reasoning) leads in reasoning capabilities with a 79.0% GPQA score, excelling at complex analytical tasks and problem-solving.
Doubao Seed Code achieves a 47.4 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 235B A22B 2507 (Reasoning)
- Complex reasoning tasks
- Research & analysis
Doubao Seed Code
- Cost-sensitive applications
- High-volume processing
- Code generation
- Software development
Try Models for Free
Try Qwen3 235B A22B 2507 (Reasoning) for FREE
No credit card or account required.
Try Doubao Seed Code for FREE
No credit card or account required.
Cost Calculator
Qwen3 235B A22B 2507 (Reasoning) Doubao Seed Code Costs are estimates based on API pricing. Actual costs may vary based on caching, batch processing, and volume discounts.
