Claude Haiku 5.5
On this page
| Developer | Anthropic |
|---|---|
| Released | October 7, 2026 |
| Status | Current |
| Inputs | Text, Images |
| Context window | 1M tokens |
| API price (input) | $0.10 / 1M tokens |
| API price (output) | $0.50 / 1M tokens |
| Open weights | No |
| Access | Claude apps, Claude API, Amazon Bedrock, Google Cloud, Microsoft Foundry, Claude Platform on AWS |
Specs and list prices last verified October 11, 2026. Prices change often; check the provider before you buy.
Strengths
- Fastest model in Anthropic's lineup
- From $0.10 / $0.50 per million tokens
- 1M-token context and 128K output
- Adaptive thinking with adjustable effort
Best for
- Classification
- extraction and routing at scale
- Sub-agents inside larger agent systems
- Latency-sensitive features in apps
Claude Haiku 5.5 is the smallest and fastest model in Anthropic’s Claude 5 lineup. Released on October 7, 2026, it is built, in Anthropic’s words, “for high-volume, latency-sensitive tasks such as classification, extraction, and routing”, plus sub-agent work inside larger agent systems.
Key specs
- Context window: 1 million tokens; up to 128,000 output tokens (up to 300,000 on the Batch API with a beta header)
- Thinking: adaptive, with “medium” default effort
- Inputs: text and images; output is text
- Knowledge cutoff: June 2026
- Availability: Claude API, Amazon Bedrock, Google Cloud, Microsoft Foundry and Claude Platform on AWS
Pricing is tiered by prompt length
| Prompt size | Input / 1M tokens | Cache read | Output / 1M tokens |
|---|---|---|---|
| Up to 100K tokens | $0.10 | $0.01 | $0.50 |
| Over 100K tokens | $0.50 | $0.05 | $2.50 |
Batch API requests are 50% off. Our tables list the up-to-100K price. Haiku 5.5 uses Anthropic’s newer tokenizer, which Anthropic says counts the same text as about 30% more tokens than Claude Haiku 4.5 did, so compare real costs on your own prompts.
Where it fits in the Claude family
| Model | Price (in / out per 1M) | Anthropic’s description |
|---|---|---|
| Claude Fable 5.1 | $10 / $50 | Demanding reasoning and long-horizon agentic work |
| Claude Opus 5.5 | $4 / $20 | Long-running agentic coding and knowledge work |
| Claude Sonnet 5.5 | $2 / $10 | Best combination of speed and intelligence |
| Claude Haiku 5.5 | From $0.10 / $0.50 | High-volume, latency-sensitive tasks |
Competitors
Haiku 5.5’s closest rival is OpenAI’s GPT-6 Luna, which has the same $0.10/$0.50 list price. See GPT-6 Luna vs Claude Haiku 5.5. Google’s Gemini 3.8 Flash is a pricier option with audio and video input.
Frequently asked questions
How much does Claude Haiku 5.5 cost?
$0.10 per million input tokens and $0.50 per million output tokens for prompts up to 100,000 tokens. Longer prompts cost $0.50 input and $2.50 output. Cache reads start at $0.01 per million tokens, and Batch API requests are 50% off.
Should I use Claude Haiku 5.5 or Claude Sonnet 5.5?
Anthropic positions Haiku 5.5 for high-volume, latency-sensitive tasks such as classification, extraction and routing, and Sonnet 5.5 as the best combination of speed and intelligence. Sonnet costs $2/$10 per million tokens, 20 times Haiku's base price.
Sources
- Claude Haiku 5.5 (model overview) — Anthropic
- Claude models overview — Anthropic
Written by
Luka Kušec · Editor
Editor of SI.info. Writes about Super Intelligence, technology policy and the people building frontier models.