Claude Haiku 5.5 API pricing
As of October 16, 2026, Claude Haiku 5.5 costs $0.10 per million input tokens and $0.50 per million output tokens on the Anthropic API. Cached input is $0.01 per million.
New Oct 7
It is the cheapest of 16 Claude models tracked here, by blended price. The closest Budget-tier model from another provider is GPT-6 Luna (OpenAI) at $0.10 / $0.50, the same blended price. Same list price as GPT-6 Luna. First listed in the LiteLLM price data on October 7, 2026. Prompts longer than 100K tokens are billed at $0.50 input / $2.50 output.
Input / 1M tokens$0.10
Output / 1M tokens$0.50
Context window1M
Max output128K
Every published rate
| Rate | Per 1M tokens |
|---|
| Input (standard) | $0.10 |
| Output (standard) | $0.50 |
| Cached input (read) | $0.01 |
| Cache write | $0.125 |
| Batch input | $0.05 |
| Batch output | $0.25 |
| Input, prompts over 100K tokens | $0.50 |
| Output, prompts over 100K tokens | $2.50 |
What it costs in practice
| Task | Tokens | Per request | Per 1,000 requests |
|---|
| Chatbot reply | 1,000 in / 500 out | $0.00035 | $0.35 |
| Summarize a long document | 20,000 in / 1,000 out | $0.0025 | $2.50 |
| Coding agent step | 50,000 in / 2,000 out | $0.006 | $6.00 |
Standard rates, no caching or batch discount; prompts over 100K tokens are billed at the higher long-prompt rate. Compare your own numbers across all models →
Price history
Dates are when each change first appeared in the LiteLLM price data, usually within days of the provider's announcement.
Details
| Provider | Anthropic |
| Price tier | Budget |
| Features | Vision, Reasoning, Prompt caching, Tool use, PDF input, Web search |
| API model names | claude-haiku-5-5 |
| Official pricing | platform.claude.com |
Models at a similar price
Other Anthropic models