Model Comparison
Magistral Medium 1.2
vs. Small (Feb '24)
Comparing 2 AI models · 12 benchmarks · Mistral
Recommended Pick
Mi Magistral Medium 1.2 8 metric wins Strongest on: Value, Reasoning, Intelligence
Best Value
Magistral Medium 1.2
100.0 value score
54.9 reasoning / $2.75/1M
Lowest Price
Small (Feb '24)
$1.00/1M input price
Best Reasoning
Magistral Medium 1.2
54.9 reasoning score
Blends available reasoning benchmarks
Best for Coding
Magistral Medium 1.2
21.7 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
Magistral Medium 1.2 has the strongest quality-to-price mix at 100.0 out of 100 value points.
Price gap
Small (Feb '24) is 2.0x cheaper on input tokens than Magistral Medium 1.2.
Speed gap
Small (Feb '24) generates about 3.8x as many tokens per second as Magistral Medium 1.2.
Reasoning gap
Magistral Medium 1.2 leads Small (Feb '24) by 34.8 points on reasoning.
Top-pick rationale
Magistral Medium 1.2 wins 8 measurable categories, including Value, Reasoning, Intelligence, GPQA.
Response Face-Off
Run one prompt through the selected models and compare response quality with live speed and cost context.
Magistral Medium 1.2
Mistral
TTFT
—
Time
—
tok/s
—
Tokens
—
Cost
—
Small (Feb '24)
Mistral
TTFT
—
Time
—
tok/s
—
Tokens
—
Cost
—
Which answer was more useful?
Full Comparison
| Metric | Top Pick Mi Magistral Medium 1.2 | Mi Small (Feb '24) |
|---|---|---|
| Pricing per 1M tokens | ||
| Input Cost | $2.00/1M | $1.00/1M |
| Output Cost | $5.00/1M | $3.00/1M |
| Blended (3:1) | $2.75/1M | $1.50/1M |
| Specifications | ||
| Organization | Mistral | Mistral |
| Release Date | Sep 18, 2025 | Feb 26, 2024 |
| Performance & Speed | ||
| Throughput | 39.9 tok/s | 153.7 tok/s |
| TTFT | 522ms | 464ms |
| Latency | 50593ms | 464ms |
| Composite Indices | ||
| Value Score | 100.0 | 67.1 |
| Reasoning Score | 54.9 | 20.1 |
| Intelligence | 27.1 | 9.0 |
| Coding | 21.7 | — |
| Math | 82.0 | — |
| Standard Benchmarks | ||
| GPQA | 73.9% | 30.2% |
| MMLU Pro | 81.5% | 41.9% |
| HLE | 9.6% | 4.4% |
| LiveCodeBench | 75.0% | 11.1% |
| MATH 500 | — | 56.2% |
| AIME 2025 | 82.0% | — |
| AIME (Original) | — | 0.7% |
| SciCode | 39.2% | 13.4% |
| LCR | 51.3% | — |
| IFBench | 43.0% | — |
| TAU-bench v2 | 52.0% | — |
| TerminalBench Hard | 12.9% | — |
Key Takeaways
Small (Feb '24) offers the best value at $1.00/1M, making it ideal for high-volume applications and cost-conscious projects.
Magistral Medium 1.2 has the strongest reasoning profile with a 54.9 reasoning score, combining the available reasoning-heavy benchmarks.
Magistral Medium 1.2 reaches a 21.7 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
Magistral Medium 1.2
- Complex reasoning tasks
- Research & analysis
- Code generation
- Software development
Small (Feb '24)
- Cost-sensitive applications
- High-volume processing
