Claude Haiku 4.5 is Anthropic's fastest, most cost-efficient model, matching Sonnet 4's performance on coding, computer use, and agent tasks. It offers similar performance to Sonnet 4 at one-third the cost and more than twice the speed, making it ideal for high-volume, latency-sensitive applications and multi-agent orchestration.
发布日期2025年10月15日
参数规模—
上下文长度200K
许可证Proprietary
知识截止2025年2月1日
Benchmarks
评测成绩
| 评测基准 | 类别 | 分数 | 来源 |
|---|---|---|---|
| AIME 2025 | math reasoning | 96.3 | 来源 |
| Tau2 Retail | communication reasoning | 83.2 | 来源 |
| Tau2 Telecom | communication reasoning | 83.0 | 来源 |
| MMMLU | language reasoning math general | 83.0 | 来源 |
| AIME 2025 | math reasoning | 80.7 | 来源 |
| SWE-Bench Verified | reasoning frontend_development code | 73.3 | 来源 |
| MMMU (validation) | vision multimodal reasoning general | 73.2 | 来源 |
| GPQA | reasoning general | 73.0 | 来源 |
| Tau2 Airline | reasoning communication | 63.6 | 来源 |
| OSWorld | multimodal general vision | 50.7 | 来源 |
| Cybersecurity CTFs | safety | 46.9 | 来源 |
| Terminal-Bench | reasoning code | 41.0 | 来源 |
Pricing
API 价格对比
| 服务商 | 输入价 | 输出价 | 上下文 | 吞吐(tok/s) | 延迟(s) | 函数调用 | 代码执行 | 联网搜索 |
|---|---|---|---|---|---|---|---|---|
| 302.AI | $1.00 | $5.00 | 200K | — | — | ✓ | ✗ | ✗ |
| Anthropic | $1.00 | $5.00 | 200K | — | — | ✓ | ✗ | ✗ |
| Azure | $1.00 | $5.00 | 200K | — | — | ✓ | ✗ | ✗ |
| Azure Cognitive Services | $1.00 | $5.00 | 200K | — | — | ✓ | ✗ | ✗ |
| FrogBot | $1.00 | $5.00 | 200K | — | — | ✓ | ✗ | ✗ |
| LLM Gateway | $1.00 | $5.00 | 200K | — | — | ✓ | ✗ | ✗ |
| Neon | $1.00 | $5.00 | 200K | — | — | ✓ | ✗ | ✗ |
| OpenCode Zen | $1.00 | $5.00 | 200K | — | — | ✓ | ✗ | ✗ |
| Pioneer | $1.00 | $5.00 | 200K | — | — | ✓ | ✗ | ✗ |
| Cortecs | $1.09 | $5.43 | 200K | — | — | ✓ | ✗ | ✗ |
价格单位：美元/百万 token，数据来自社区整理，仅供参考。