Model Comparison
Small (Feb '24)
vs. Qwen3 32B (Non-reasoning)
Comparing 2 AI models · 10 benchmarks · Mistral, Alibaba
Recommended Pick
Al Qwen3 32B (Non-reasoning) 12 metric wins Strongest on: Value, Input price, Output price
Best Value
Qwen3 32B (Non-reasoning)
100.0 value score
32.7 reasoning / $0.26/1M
Lowest Price
Qwen3 32B (Non-reasoning)
$0.15/1M input price
Best Reasoning
Qwen3 32B (Non-reasoning)
32.7 reasoning score
Blends available reasoning benchmarks
Composite Indices
Higher is better; speed and price are normalized
Standard Benchmarks
Only benchmarks with data are shown
Differences That Matter
Best value
Qwen3 32B (Non-reasoning) has the strongest quality-to-price mix at 100.0 out of 100 value points.
Price gap
Qwen3 32B (Non-reasoning) is 6.7x cheaper on input tokens than Small (Feb '24).
Speed gap
Small (Feb '24) generates about 1.9x as many tokens per second as Qwen3 32B (Non-reasoning).
Reasoning gap
Qwen3 32B (Non-reasoning) leads Small (Feb '24) by 12.6 points on reasoning.
Top-pick rationale
Qwen3 32B (Non-reasoning) wins 12 measurable categories, including Value, Input price, Output price, Blended price.
Response Face-Off
Run one prompt through the selected models and compare response quality with live speed and cost context.
Small (Feb '24)
Mistral
TTFT
—
Time
—
tok/s
—
Tokens
—
Cost
—
Qwen3 32B (Non-reasoning)
Alibaba
TTFT
—
Time
—
tok/s
—
Tokens
—
Cost
—
Which answer was more useful?
Full Comparison
| Metric | Mi Small (Feb '24) | Top Pick Al Qwen3 32B (Non-reasoning) |
|---|---|---|
| Pricing per 1M tokens | ||
| Input Cost | $1.00/1M | $0.15/1M |
| Output Cost | $3.00/1M | $0.59/1M |
| Blended (3:1) | $1.50/1M | $0.26/1M |
| Specifications | ||
| Organization | Mistral | Alibaba |
| Release Date | Feb 26, 2024 | Apr 28, 2025 |
| Performance & Speed | ||
| Throughput | 170.2 tok/s | 90.1 tok/s |
| TTFT | 520ms | 1140ms |
| Latency | 520ms | 1140ms |
| Composite Indices | ||
| Value Score | 10.7 | 100.0 |
| Reasoning Score | 20.1 | 32.7 |
| Intelligence | 9.0 | 14.5 |
| Math | — | 19.7 |
| Standard Benchmarks | ||
| GPQA | 30.2% | 53.5% |
| MMLU Pro | 41.9% | 72.7% |
| HLE | 4.4% | 4.3% |
| LiveCodeBench | 11.1% | 28.8% |
| MATH 500 | 56.2% | 86.9% |
| AIME 2025 | — | 19.7% |
| AIME (Original) | 0.7% | 30.3% |
| SciCode | 13.4% | 28.0% |
| LCR | — | 0.0% |
| IFBench | — | 31.5% |
Key Takeaways
Qwen3 32B (Non-reasoning) offers the best value at $0.15/1M, making it ideal for high-volume applications and cost-conscious projects.
Qwen3 32B (Non-reasoning) has the strongest reasoning profile with a 32.7 reasoning score, combining the available reasoning-heavy benchmarks.
All models support context windows of ∞+ tokens, suitable for processing lengthy documents and maintaining extended conversations.
When to Choose Each Model
Small (Feb '24)
- General-purpose AI
- Versatile applications
Qwen3 32B (Non-reasoning)
- Cost-sensitive applications
- High-volume processing
- Complex reasoning tasks
- Research & analysis
