Model Comparison
Claude 3.7 Sonnet (Reasoning)
vs. Gemini 2.5 Flash Preview (Sep '25) (Non-reasoning)
Comparing 2 AI models · 12 benchmarks · Anthropic, Google
Recommended Pick
An Claude 3.7 Sonnet (Reasoning) 11 metric wins Strongest on: Reasoning, Intelligence, Coding
Lowest Price
Claude 3.7 Sonnet (Reasoning)
$0.00/1M input price
Best Reasoning
Claude 3.7 Sonnet (Reasoning)
54.0 reasoning score
Blends available reasoning benchmarks
Best for Coding
Claude 3.7 Sonnet (Reasoning)
27.6 coding index
Composite Indices
Higher is better; speed and price are normalized
Standard Benchmarks
Only benchmarks with data are shown
Differences That Matter
Reasoning gap
Claude 3.7 Sonnet (Reasoning) leads Gemini 2.5 Flash Preview (Sep '25) (Non-reasoning) by 9.3 points on reasoning.
Coding gap
Claude 3.7 Sonnet (Reasoning) leads Gemini 2.5 Flash Preview (Sep '25) (Non-reasoning) by 5.5 points on coding.
Top-pick rationale
Claude 3.7 Sonnet (Reasoning) wins 11 measurable categories, including Reasoning, Intelligence, Coding, GPQA.
Response Face-Off
Run one prompt through the selected models and compare response quality with live speed and cost context.
Claude 3.7 Sonnet (Reasoning)
Anthropic
TTFT
—
Time
—
tok/s
—
Tokens
—
Cost
—
Gemini 2.5 Flash Preview (Sep '25) (Non-reasoning)
TTFT
—
Time
—
tok/s
—
Tokens
—
Cost
—
Which answer was more useful?
Full Comparison
| Metric | Top Pick An Claude 3.7 Sonnet (Reasoning) | Go Gemini 2.5 Flash Preview (Sep '25) (Non-reasoning) |
|---|---|---|
| Pricing per 1M tokens | ||
| Input Cost | $0.00/1M | $0.00/1M |
| Output Cost | $0.00/1M | $0.00/1M |
| Specifications | ||
| Organization | Anthropic | |
| Release Date | Feb 24, 2025 | Sep 25, 2025 |
| Performance & Speed | ||
| Throughput | — | — |
| TTFT | — | — |
| Latency | — | — |
| Composite Indices | ||
| Reasoning Score | 54.0 | 44.7 |
| Intelligence | 34.7 | 25.7 |
| Coding | 27.6 | 22.1 |
| Math | 56.3 | 56.7 |
| Standard Benchmarks | ||
| GPQA | 77.2% | 76.6% |
| MMLU Pro | 83.7% | 83.6% |
| HLE | 10.3% | 7.8% |
| LiveCodeBench | 47.3% | 62.5% |
| MATH 500 | 94.7% | — |
| AIME 2025 | 56.3% | 56.7% |
| AIME (Original) | 48.7% | — |
| SciCode | 40.3% | 37.5% |
| LCR | 60.7% | 56.7% |
| IFBench | 48.3% | 43.5% |
| TAU-bench v2 | 54.7% | 28.4% |
| TerminalBench Hard | 21.2% | 14.4% |
Key Takeaways
Claude 3.7 Sonnet (Reasoning) offers the best value at $0.00/1M, making it ideal for high-volume applications and cost-conscious projects.
Claude 3.7 Sonnet (Reasoning) has the strongest reasoning profile with a 54.0 reasoning score, combining the available reasoning-heavy benchmarks.
Claude 3.7 Sonnet (Reasoning) reaches a 27.6 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
Claude 3.7 Sonnet (Reasoning)
- Cost-sensitive applications
- High-volume processing
- Complex reasoning tasks
- Research & analysis
- Code generation
- Software development
Gemini 2.5 Flash Preview (Sep '25) (Non-reasoning)
- General-purpose AI
- Versatile applications
