Claude Haiku 5.5: Anthropic's cheapest small model gets a big upgrade

Anthropic released Claude Haiku 5.5 on 7 October. It's aimed at high-volume work such as summaries, compaction, database queries and classification, and Anthropic also pitches it as a subagent next to Opus 5.5 and Sonnet 5.5 for coding. It's the first Haiku with an adjustable effort setting, and it has a 1M-token context window with up to 128k output tokens.

Pricing: where the 75% comes from

• Prompts up to 100,000 tokens: $0.10 per million input tokens and $0.50 per million output, 90% below Haiku 4.5's $1 / $5.
• Prompts over 100,000 tokens: $0.50 input and $2.50 output, 50% below Haiku 4.5.
• Cache hits are $0.01 per million (short prompts) and $0.05 (long prompts).

Anthropic says about 90% of Haiku 4.5 requests fell under the 100k threshold. That, adjusted for a new tokenizer that uses slightly more tokens per task, is how it arrives at "around 75% less to run" on average.

Benchmarks (Anthropic's numbers)

• OSWorld 2.1 (offline subset): 72.4%, vs 15.7% for Haiku 4.5 and 48.9% for GPT-6 Luna.
• Terminal-Bench 4.0: 39.2%, vs 0.0% for Haiku 4.5 and 16.4% for GPT-6 Luna.
• GDPval-AA v2.1: 1620 Elo, vs 735 for Haiku 4.5 and 1437 for GPT-6 Luna.
• Sonnet 5.5 is still well ahead: 83.9% on OSWorld 2.1 and 70.6% on Terminal-Bench 4.0.

Where you can use it

The model ID is claude-haiku-5-5. It's on the Claude API, Amazon Bedrock, Claude Platform on AWS, Google Cloud and Microsoft Foundry. In Claude Code, version 2.1.293 made it the default Haiku model on the Anthropic API.