Model Comparison
Medium 3
vs. Step3 VL 10B
Comparing 2 AI models · 12 benchmarks · Mistral, StepFun
Recommended Pick
St Step3 VL 10B 9 metric wins Strongest on: Input price, Output price, TTFT
Lowest Price
Step3 VL 10B
$0.00/1M input price
Best Reasoning
Medium 3
39.5 reasoning score
Blends available reasoning benchmarks
Best for Coding
Step3 VL 10B
13.9 coding index
Composite Indices
Higher is better; speed and price are normalized
Standard Benchmarks
Only benchmarks with data are shown
Differences That Matter
Price gap
Step3 VL 10B is ∞x cheaper on input tokens than Medium 3.
Reasoning gap
Medium 3 leads Step3 VL 10B by 7.9 points on reasoning.
Coding gap
Step3 VL 10B leads Medium 3 by 0.3 points on coding.
Top-pick rationale
Step3 VL 10B wins 9 measurable categories, including Input price, Output price, TTFT, Latency.
Response Face-Off
Run one prompt through the selected models and compare response quality with live speed and cost context.
Medium 3
Mistral
TTFT
—
Time
—
tok/s
—
Tokens
—
Cost
—
Step3 VL 10B
StepFun
TTFT
—
Time
—
tok/s
—
Tokens
—
Cost
—
Which answer was more useful?
Full Comparison
| Metric | Mi Medium 3 | Top Pick St Step3 VL 10B |
|---|---|---|
| Pricing per 1M tokens | ||
| Input Cost | $0.40/1M | $0.00/1M |
| Output Cost | $2.00/1M | $0.00/1M |
| Blended (3:1) | $0.80/1M | — |
| Specifications | ||
| Organization | Mistral | StepFun |
| Release Date | May 7, 2025 | Jan 20, 2026 |
| Performance & Speed | ||
| Throughput | 46.1 tok/s | — |
| TTFT | 533ms | — |
| Latency | 533ms | — |
| Composite Indices | ||
| Value Score | 100.0 | — |
| Reasoning Score | 39.5 | 31.6 |
| Intelligence | 18.8 | 15.5 |
| Coding | 13.6 | 13.9 |
| Math | 30.3 | — |
| Standard Benchmarks | ||
| GPQA | 57.8% | 69.0% |
| MMLU Pro | 76.0% | — |
| HLE | 4.3% | 10.2% |
| LiveCodeBench | 40.0% | — |
| MATH 500 | 90.7% | — |
| AIME 2025 | 30.3% | — |
| AIME (Original) | 44.0% | — |
| SciCode | 33.1% | 31.1% |
| LCR | 28.0% | 0.0% |
| IFBench | 39.3% | 50.2% |
| TAU-bench v2 | 24.3% | 16.1% |
| TerminalBench Hard | 3.8% | 5.3% |
Key Takeaways
Step3 VL 10B offers the best value at $0.00/1M, making it ideal for high-volume applications and cost-conscious projects.
Medium 3 has the strongest reasoning profile with a 39.5 reasoning score, combining the available reasoning-heavy benchmarks.
Step3 VL 10B reaches a 13.9 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
Medium 3
- Complex reasoning tasks
- Research & analysis
Step3 VL 10B
- Cost-sensitive applications
- High-volume processing
- Code generation
- Software development
