Claude Haiku 5.5 is Anthropic's smallest and fastest model in the Claude 5.5 family, built for teams that need high-volume, cheapest, low-latency inference. Released on October 7, 2026, it targets the repetitive but important jobs: document summarization, classification, routing, and the helper sub-agents that feed larger models such as Sonnet 5.5 and Opus 5.5.
Core Features
- Prompt-length tiered pricing starting at $0.10 per million input tokens and $0.50 per million output tokens for prompts under 100,000 tokens, about 90% cheaper than Haiku 4.5.
- 1.05 million token context window with up to 128,000 output tokens, so it can summarize long threads or codebases in one pass.
- First Haiku with adjustable effort controls and adaptive thinking, default medium, letting you trade latency for accuracy per call.
- Multimodal input covering text and images, plus beta computer-use and browser-use tools through the Python and TypeScript SDKs.
- Available on the Claude API, Amazon Bedrock, Google Cloud Vertex AI, and Microsoft Foundry.
Use Cases
- High-volume content tagging, sentiment sorting, and database queries where cost per call dominates.
- Live customer-support triage and routing inside larger agent systems.
- Sub-agent retrieval for coding copilots, as financial AI firm Rogo does for revenue lookups.
- Context compaction and summarization that keep bigger models within token budgets.
Pricing
Pay-as-you-go at $0.10 in and $0.50 out per 1M tokens for prompts up to 100,000 tokens; above that it rises to $0.50 and $2.50. The Batch API adds 50% off, and cache reads cost $0.01 per 1M. No subscription is required.
Our Take
Best for cost-sensitive, high-throughput pipelines where a 90% price cut beats frontier accuracy. Trade-off: hard reasoning and complex coding still belong to Sonnet 5.5 or Opus 5.5, and the over-100K-token tier stings fivefold. Compare it with other models in our AI Engine/Model directory.
