Claude 4 Sonnet (Reasoning) vs Claude 4.5 Sonnet (Reasoning)
Comparing 2 AI models · 6 benchmarks · Anthropic
Most Affordable
An
Claude 4 Sonnet (Reasoning)
$3.00/1M
Highest Intelligence
An
Claude 4.5 Sonnet (Reasoning)
83.4% GPQA
Best for Coding
An
Claude 4.5 Sonnet (Reasoning)
37.1 Coding Index
Price Difference
1.0x
input cost range
Composite Indices
Intelligence, Coding, Math
Standard Benchmarks
Academic and industry benchmarks
Benchmark Winners
6 tests
An
Claude 4 Sonnet (Reasoning)
1
- MATH 500
An
Claude 4.5 Sonnet (Reasoning)
5
- GPQA
- MMLU Pro
- HLE
- LiveCodeBench
- AIME 2025
| Metric | An Claude 4 Sonnet (Reasoning) | An Claude 4.5 Sonnet (Reasoning) |
|---|---|---|
| Pricing Per 1M tokens | ||
| Input Cost | $3.00/1M | $3.00/1M |
| Output Cost | $15.00/1M | $15.00/1M |
| Blended Cost 3:1 input/output ratio | $6.00/1M | $6.00/1M |
| Specifications | ||
| Organization Model creator | Anthropic | Anthropic |
| Release Date Launch date | May 22, 2025 | Sep 29, 2025 |
| Performance & Speed | ||
| Throughput Output speed | 69.9 tok/s | 80.6 tok/s |
| Time to First Token (TTFT) Initial response delay | 753ms | 1666ms |
| Latency Time to first answer token | 29381ms | 26482ms |
| Composite Indices | ||
| Intelligence Index Overall reasoning capability | 38.4 | 42.4 |
| Coding Index Programming ability | 33.2 | 37.1 |
| Math Index Mathematical reasoning | 74.3 | 88.0 |
| Standard Benchmarks | ||
| GPQA Graduate-level reasoning | 77.7% | 83.4% |
| MMLU Pro Advanced knowledge | 84.2% | 87.5% |
| HLE Hard language evaluation | 9.6% | 17.3% |
| LiveCodeBench Real-world coding tasks | 65.5% | 71.4% |
| MATH 500 Mathematical problems | 99.1% | — |
| AIME 2025 Advanced math competition | 74.3% | 88.0% |
| AIME (Original) Math olympiad problems | 77.3% | — |
| SciCode Scientific code generation | 40.0% | 44.7% |
| LCR Code review capability | 64.7% | 65.7% |
| IFBench Instruction-following | 54.7% | 57.3% |
| TAU-bench v2 Tool use & agentic tasks | 64.6% | 78.1% |
| TerminalBench Hard CLI command generation | 29.8% | 33.3% |
Key Takeaways
Claude 4 Sonnet (Reasoning) offers the best value at $3.00/1M, making it ideal for high-volume applications and cost-conscious projects.
Claude 4.5 Sonnet (Reasoning) leads in reasoning capabilities with a 83.4% GPQA score, excelling at complex analytical tasks and problem-solving.
Claude 4.5 Sonnet (Reasoning) achieves a 37.1