Model Comparison
Qwen3.5 397B A17B (Reasoning)
vs. Step 3.5 Flash 2603
Comparing 2 AI models · 7 benchmarks · Alibaba, StepFun
Recommended Pick
Al Qwen3.5 397B A17B (Reasoning) 10 metric wins Strongest on: Reasoning, Intelligence, Coding
Best Value
Step 3.5 Flash 2603
100.0 value score
47.9 reasoning / $0.15/1M
Lowest Price
Step 3.5 Flash 2603
$0.10/1M input price
Best Reasoning
Qwen3.5 397B A17B (Reasoning)
53.9 reasoning score
Blends available reasoning benchmarks
Best for Coding
Qwen3.5 397B A17B (Reasoning)
41.3 coding index
Composite Indices
Higher is better; speed and price are normalized
Standard Benchmarks
Only benchmarks with data are shown
Differences That Matter
Best value
Step 3.5 Flash 2603 has the strongest quality-to-price mix at 100.0 out of 100 value points.
Price gap
Step 3.5 Flash 2603 is 6.0x cheaper on input tokens than Qwen3.5 397B A17B (Reasoning).
Speed gap
Step 3.5 Flash 2603 generates about 3.7x as many tokens per second as Qwen3.5 397B A17B (Reasoning).
Reasoning gap
Qwen3.5 397B A17B (Reasoning) leads Step 3.5 Flash 2603 by 6.0 points on reasoning.
Coding gap
Qwen3.5 397B A17B (Reasoning) leads Step 3.5 Flash 2603 by 6.7 points on coding.
Response Face-Off
Run one prompt through the selected models and compare response quality with live speed and cost context.
Qwen3.5 397B A17B (Reasoning)
Alibaba
TTFT
—
Time
—
tok/s
—
Tokens
—
Cost
—
Step 3.5 Flash 2603
StepFun
TTFT
—
Time
—
tok/s
—
Tokens
—
Cost
—
Which answer was more useful?
Full Comparison
| Metric | Top Pick Al Qwen3.5 397B A17B (Reasoning) | St Step 3.5 Flash 2603 |
|---|---|---|
| Pricing per 1M tokens | ||
| Input Cost | $0.60/1M | $0.10/1M |
| Output Cost | $3.60/1M | $0.30/1M |
| Blended (3:1) | $1.35/1M | $0.15/1M |
| Specifications | ||
| Organization | Alibaba | StepFun |
| Release Date | Feb 16, 2026 | Apr 2, 2026 |
| Performance & Speed | ||
| Throughput | 52.4 tok/s | 195.9 tok/s |
| TTFT | 1885ms | 1315ms |
| Latency | 62706ms | 11525ms |
| Composite Indices | ||
| Value Score | 12.5 | 100.0 |
| Reasoning Score | 53.9 | 47.9 |
| Intelligence | 45.0 | 38.5 |
| Coding | 41.3 | 34.6 |
| Standard Benchmarks | ||
| GPQA | 89.3% | 82.6% |
| HLE | 27.3% | 22.6% |
| SciCode | 42.0% | 38.5% |
| LCR | 65.7% | 54.3% |
| IFBench | 78.8% | 66.5% |
| TAU-bench v2 | 95.6% | 87.4% |
| TerminalBench Hard | 40.9% | 32.6% |
Key Takeaways
Step 3.5 Flash 2603 offers the best value at $0.10/1M, making it ideal for high-volume applications and cost-conscious projects.
Qwen3.5 397B A17B (Reasoning) has the strongest reasoning profile with a 53.9 reasoning score, combining the available reasoning-heavy benchmarks.
Qwen3.5 397B A17B (Reasoning) reaches a 41.3 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
Qwen3.5 397B A17B (Reasoning)
- Complex reasoning tasks
- Research & analysis
- Code generation
- Software development
Step 3.5 Flash 2603
- Cost-sensitive applications
- High-volume processing
