Anthropic drops Claude Haiku 5.5: The race to $0.10 is finally here
After nearly a year, Anthropic finally decided that being cheap is a virtue. The new Claude Haiku 5.5 is here to handle your grunt work without burning a hole in your wallet, assuming you care about pennies while the AI overlords plot our demise.
Anthropic has officially launched Claude Haiku 5.5, the latest iteration in its model lineup and the first to feature a user-tunable effort level. By moving away from the fixed output style of its predecessor, the model now lets users choose between raw speed and extended, thoughtful reasoning.
This release is clearly aimed at high-frequency tasks like data extraction, summarization, and classification, where using the heavy-duty Claude Sonnet or Opus would be like hiring a surgeon to change a lightbulb. The pricing structure is the real headline here, with costs starting at $0.10 per million input tokens and $0.50 per million output tokens for standard requests up to 100k tokens. This puts the model in direct price-parity with the GPT-6 Luna.
For workloads requiring longer context windows, the pricing scales up to $0.50 and $2.50 respectively. Anthropic claims this is roughly 90% cheaper for the bulk of typical traffic compared to Claude Haiku 4.5, effectively making it the go-to utility player for automated agents and browser-based tasks. The average cost reduction across all use cases sits at about 75%.
It is truly heartwarming to see giant tech companies fighting a desperate price war to see who can provide the cheapest cognitive slave labor. The inevitable commoditization of intelligence suggests that soon, the cost of generating a thousand useless emails will be lower than the cost of a cup of coffee.
Comments
Help shape the next version: Add context or suggest a correction. AI review can add points toward a rewrite. Reviews and updates may take time; a full meter does not guarantee a new version.