Anthropic releases Claude Haiku 5.5 at $0.10 per million input tokens, a tenth of Haiku 4.5's price for prompts up to 100,000 tokens

Anthropic released Claude Haiku 5.5 on 7 October, and calls it its cheapest and fastest small model yet. The launch page lists the price for prompts up to 100,000 tokens as $0.10 per million input tokens and $0.50 per million output tokens. Haiku 4.5 cost $1 and $5. Longer prompts cost $0.50 and $2.50.

Anthropic says Haiku 5.5 is “priced 90% lower than Claude Haiku 4.5 for requests up to 100,000 tokens”, and about 50% lower above that. Averaged over real use, it says, “it now costs around 75% less to run.” The page says about 90% of requests to Haiku 4.5 were the shorter kind.

It is the first Haiku with an adjustable effort setting. Anthropic aims it at high-volume work such as summaries, classification, database queries and acting as a subagent, and still points developers to Sonnet 5.5 and Opus 5.5 for complex agentic coding. The company's own benchmark table shows large gains over Haiku 4.5: 72.4% against 15.7% on an offline subset of OSWorld 2.1, and 39.2% against 0.0% on Terminal-Bench 4.0. The page says its cyber safeguards are stricter than Haiku 4.5's but looser than Sonnet 5.5's. It is on the Claude API and the major clouds.

• Confirmed Released 7 October at $0.10/$0.50 per million tokens for prompts up to 100,000 tokens, and $0.50/$2.50 above. Anthropic
• Claimed Anthropic says that on average “it now costs around 75% less to run.” Anthropic
• Claimed Anthropic's table shows 72.4% against Haiku 4.5's 15.7% on an OSWorld 2.1 offline subset. Anthropic