Model Comparison
Claude 3.5 Sonnet (June '24)
vs. GPT-5 (high)
Comparing 2 AI models · 12 benchmarks · Anthropic, OpenAI
Recommended Pick
Op GPT-5 (high) 14 metric wins Strongest on: Value, Input price, Output price
Best Value
GPT-5 (high)
100.0 value score
77.2 reasoning / $3.44/1M
Lowest Price
GPT-5 (high)
$1.25/1M input price
Best Reasoning
GPT-5 (high)
77.2 reasoning score
Blends available reasoning benchmarks
Best for Coding
GPT-5 (high)
36.0 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
GPT-5 (high) has the strongest quality-to-price mix at 100.0 out of 100 value points.
Price gap
GPT-5 (high) is 3.0x cheaper on input tokens than Claude 3.5 Sonnet (June '24).
Reasoning gap
GPT-5 (high) leads Claude 3.5 Sonnet (June '24) by 46.6 points on reasoning.
Coding gap
GPT-5 (high) leads Claude 3.5 Sonnet (June '24) by 10.0 points on coding.
Top-pick rationale
GPT-5 (high) wins 14 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.
Claude 3.5 Sonnet (June '24)
Anthropic
TTFT
—
Time
—
tok/s
—
Tokens
—
Cost
—
GPT-5 (high)
OpenAI
TTFT
—
Time
—
tok/s
—
Tokens
—
Cost
—
Which answer was more useful?
Full Comparison
| Metric | An Claude 3.5 Sonnet (June '24) | Top Pick Op GPT-5 (high) |
|---|---|---|
| Pricing per 1M tokens | ||
| Input Cost | $3.75/1M | $1.25/1M |
| Output Cost | $15.00/1M | $10.00/1M |
| Blended (3:1) | $6.56/1M | $3.44/1M |
| Specifications | ||
| Organization | Anthropic | OpenAI |
| Release Date | Jun 21, 2024 | Aug 7, 2025 |
| Performance & Speed | ||
| Throughput | — | 103.2 tok/s |
| TTFT | — | 87413ms |
| Latency | — | 87413ms |
| Composite Indices | ||
| Value Score | 20.8 | 100.0 |
| Reasoning Score | 30.6 | 77.2 |
| Intelligence | 14.2 | 44.6 |
| Coding | 26.0 | 36.0 |
| Math | — | 94.3 |
| Standard Benchmarks | ||
| GPQA | 56.0% | 85.4% |
| MMLU Pro | 75.1% | 87.1% |
| HLE | 3.7% | 26.5% |
| LiveCodeBench | — | 84.6% |
| MATH 500 | 69.5% | 99.4% |
| AIME 2025 | — | 94.3% |
| AIME (Original) | 9.7% | 95.7% |
| SciCode | 31.6% | 42.9% |
| LCR | — | 75.6% |
| IFBench | — | 73.1% |
| TAU-bench v2 | — | 84.8% |
| TerminalBench Hard | — | 32.6% |
Key Takeaways
GPT-5 (high) offers the best value at $1.25/1M, making it ideal for high-volume applications and cost-conscious projects.
GPT-5 (high) has the strongest reasoning profile with a 77.2 reasoning score, combining the available reasoning-heavy benchmarks.
GPT-5 (high) reaches a 36.0 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 3.5 Sonnet (June '24)
- General-purpose AI
- Versatile applications
GPT-5 (high)
- Cost-sensitive applications
- High-volume processing
- Complex reasoning tasks
- Research & analysis
- Code generation
- Software development
