Claude Haiku 5.5
·
A smaller Claude model for classification, summaries, routing and subagent work. Haiku now supports adjustable effort and adaptive thinking for tasks with different latency budgets.
- Reviewed tasks
- Classification and routing, Writing, Coding, Computer use
- Reviewed modalities
- Text, Images
- Deployment paths
- Hosted API
- Focus / inputs
- Text and images → text
- Availability / restrictions
- Claude Platform and supported cloud platforms. The new tokenizer and unsupported sampling settings need checking before migration.
- Licence category
- Provider terms
- Licence / terms
- Proprietary; provider terms apply.
Variants, limits and list prices
Values are copied from the linked first-party pages as stated, with the date each group was checked. Unknown means the pages checked do not state the value. Prices are list prices, not a quote or a ranking.
| Field | Claude Haiku 5.5 |
|---|---|
| API identifier | claude-haiku-5-5 |
| Released | |
| Status and retirement | Retirement not before . |
| Lifecycle checked | Checked |
| Context window | 1M tokens |
| Max input | Unknown |
| Max output | 128K tokens; 300K through the Batch API (beta) |
| Rate limits | Unknown |
| Limit notes | Unknown |
| Limits checked | Checked |
| Base input price | 0.1 USD/MTok (Prompts up to 100K tokens) |
| Base output price | 0.5 USD/MTok (Prompts up to 100K tokens) |
| Cache-read price | 0.01 USD/MTok (Prompts up to 100K tokens) |
| Price conditions | Prompts up to 100K tokens: input 0.1 USD/MTok, output 0.5 USD/MTok, cache read 0.01 USD/MTok, cache write 0.125 USD/MTok (5 minutes), cache write 0.2 USD/MTok (1 hour). Prompts over 100K tokens: input 0.5 USD/MTok, output 2.5 USD/MTok, cache read 0.05 USD/MTok, cache write 0.625 USD/MTok (5 minutes), cache write 1 USD/MTok (1 hour). 50% off input and output with the Batch API. |
| Prices checked | Checked |
| Knowledge cutoff | |
| Cutoff checked | Checked |
Parameters and weights
Unknown: the first-party pages checked publish no parameter count or weights.