Claude Haiku 4.5 vs GPT-5.4 nano
Detailed comparison of API pricing, technical specifications, and benchmark scores. Updated 2026-06-29.
Cost at a Glance
Claude Haiku 4.5 is 4.3× more expensive than GPT-5.4 nano — switching could save 76.7% on your monthly API bill.
At 5M input + 2M output tokens per month:
Choosing GPT-5.4 nano could save you $11.50/mo
Price Comparison (per 1M tokens)
| Metric | Claude Haiku 4.5 | GPT-5.4 nano |
|---|---|---|
| Input (per 1M tokens) | $1.00 | $0.20 |
| Output (per 1M tokens) | $5.00 | $1.25 |
| Cache Write (per 1M tokens) | $1.25 | $0.25 |
| Cache Read (per 1M tokens) | $0.10 | $0.02 |
| Batch Input (per 1M tokens) | $0.50 | $0.10 |
| Batch Output (per 1M tokens) | $2.50 | $0.63 |
Prices are per 1 million tokens. Actual costs depend on your monthly token volume. Use the calculator →
Specifications
| Spec | Claude Haiku 4.5 | GPT-5.4 nano |
|---|---|---|
| Context Window | 1M | 400K |
| Max Output Tokens | 64K | 128K |
| Knowledge Cutoff | 2026-03 | 2025-08 |
| Modalities | Text, Image | Text, Image |
| Prompt Caching | Yes | Yes |
| Batch API | Yes | Yes |
| Thinking / Reasoning | No | Yes |
Benchmark Scores
| Benchmark | Claude Haiku 4.5 | GPT-5.4 nano |
|---|---|---|
| SWE-bench Verified | 55.3 | 65.0 |
| MMLU-Pro | 76.8 | 74.0 |
| HumanEval | 88.2 | 85.0 |
Benchmark scores from official provider reports. Higher is better. Missing scores indicate the provider has not published that benchmark yet.
Use Cases
Claude Haiku 4.5
GPT-5.4 nano
About the Providers
Claude Haiku 4.5 is developed by Anthropic. GPT-5.4 nano is developed by OpenAI. This is a cross-provider comparison between Anthropic and OpenAI models. For the most up-to-date pricing, always check the official API documentation from each provider.