Model Comparison
Large 3
vs. Qwen3 235B A22B (Non-reasoning)
Comparing 2 AI models · 12 benchmarks · Mistral, Alibaba
Recommended Pick
Mi Large 3 14 metric wins Strongest on: Output price, Blended price, TTFT
Best Value
Qwen3 235B A22B (Non-reasoning)
100.0 value score
36.2 reasoning / $0.79/1M
Lowest Price
Qwen3 235B A22B (Non-reasoning)
$0.45/1M input price
Best Reasoning
Qwen3 235B A22B (Non-reasoning)
36.2 reasoning score
Blends available reasoning benchmarks
Best for Coding
Large 3
22.7 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
Qwen3 235B A22B (Non-reasoning) has the strongest quality-to-price mix at 100.0 out of 100 value points.
Price gap
Qwen3 235B A22B (Non-reasoning) is 1.1x cheaper on input tokens than Large 3.
Speed gap
Qwen3 235B A22B (Non-reasoning) generates about 1.1x as many tokens per second as Large 3.
Reasoning gap
Qwen3 235B A22B (Non-reasoning) leads Large 3 by 2.0 points on reasoning.
Coding gap
Large 3 leads Qwen3 235B A22B (Non-reasoning) by 8.7 points on coding.
Response Face-Off
Run one prompt through the selected models and compare response quality with live speed and cost context.
Large 3
Mistral
TTFT
—
Time
—
tok/s
—
Tokens
—
Cost
—
Qwen3 235B A22B (Non-reasoning)
Alibaba
TTFT
—
Time
—
tok/s
—
Tokens
—
Cost
—
Which answer was more useful?
Full Comparison
| Metric | Top Pick Mi Large 3 | Al Qwen3 235B A22B (Non-reasoning) |
|---|---|---|
| Pricing per 1M tokens | ||
| Input Cost | $0.50/1M | $0.45/1M |
| Output Cost | $1.50/1M | $1.80/1M |
| Blended (3:1) | $0.75/1M | $0.79/1M |
| Specifications | ||
| Organization | Mistral | Alibaba |
| Release Date | Dec 2, 2025 | Apr 28, 2025 |
| Performance & Speed | ||
| Throughput | 62.0 tok/s | 67.2 tok/s |
| TTFT | 560ms | 1167ms |
| Latency | 560ms | 1167ms |
| Composite Indices | ||
| Value Score | 99.2 | 100.0 |
| Reasoning Score | 34.2 | 36.2 |
| Intelligence | 22.8 | 17.0 |
| Coding | 22.7 | 14.0 |
| Math | 38.0 | 23.7 |
| Standard Benchmarks | ||
| GPQA | 68.0% | 61.3% |
| MMLU Pro | 80.7% | 76.2% |
| HLE | 4.1% | 4.7% |
| LiveCodeBench | 46.5% | 34.3% |
| MATH 500 | — | 90.2% |
| AIME 2025 | 38.0% | 23.7% |
| AIME (Original) | — | 32.7% |
| SciCode | 36.2% | 29.9% |
| LCR | 34.7% | 0.0% |
| IFBench | 36.2% | 36.6% |
| TAU-bench v2 | 24.6% | 27.2% |
| TerminalBench Hard | 15.9% | 6.1% |
Key Takeaways
Qwen3 235B A22B (Non-reasoning) offers the best value at $0.45/1M, making it ideal for high-volume applications and cost-conscious projects.
Qwen3 235B A22B (Non-reasoning) has the strongest reasoning profile with a 36.2 reasoning score, combining the available reasoning-heavy benchmarks.
Large 3 reaches a 22.7 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
Large 3
- Code generation
- Software development
Qwen3 235B A22B (Non-reasoning)
- Cost-sensitive applications
- High-volume processing
- Complex reasoning tasks
- Research & analysis
