Anthropic released Claude Haiku 5.5 on 7 October 2026 and says that, on average, it costs around 75% less to run than Haiku 4.5. Its pricing table lists Haiku 5.5 at $0.10 per million input tokens and $0.50 per million output tokens on prompts up to 100,000 tokens, against $1.00 and $5.00 for Haiku 4.5.
Anthropic also halved the price of Claude Sonnet 5.5’s cache reads, to $0.10 per million tokens from $0.20, which it says makes Sonnet 5.5 around 20% cheaper on most agentic work.
Business Pill · PROMPT CACHING
A one-minute explainer of prompt caching: when the same text is sent to a model again, the repeated part can be billed at a fraction of the price. It teaches the general idea only and says nothing about any company in this story.
The key insight: As we read it, Anthropic is pricing for volume work done by agents rather than for chat. Its post aims Haiku 5.5 at summaries, compaction and subagent tasks, cuts the cache reads that agents consume most, and gives subscribers API credit to build with.
On prompts over 100,000 tokens, Haiku 5.5 costs $0.50 per million input tokens and $2.50 per million output tokens, according to the same table. Cache reads cost $0.01 per million tokens up to 100,000 and $0.05 above it, against $0.10 for Haiku 4.5.
Anthropic says prompts up to 100,000 tokens make up around 90% of requests to its previous Haiku model. On our arithmetic, the per-token input and output prices on those prompts are 90% below Haiku 4.5’s; Anthropic’s own figure for the average cost to run is around 75% less.
Price per million tokens, Claude Haiku 4.5 vs Haiku 5.5 on prompts up to 100,000 tokens, from Anthropic’s pricing table of 7 October 2026: input $1.00 vs $0.10, output $5.00 vs $0.50, cache writes $1.25 vs $0.125.
What Anthropic Says It Can Do
Anthropic describes Haiku 5.5 as designed for high-volume, cost-sensitive tasks such as summaries, compactions, database queries and classification, and says it pairs with Opus 5.5 and Sonnet 5.5 as a subagent on coding work. It calls Haiku 5.5 its fastest model to date.
In Anthropic’s benchmark table, Haiku 5.5 scores 72.4% on the offline subset of OSWorld 2.1, a computer-use test, against 15.7% for Haiku 4.5, 48.9% for OpenAI’s GPT-6 Luna and 83.9% for Sonnet 5.5. On Terminal-Bench 4.0 it scores 39.2%, against 0.0% for Haiku 4.5 and 70.6% for Sonnet 5.5.
Anthropic writes that Sonnet 5.5 and Opus 5.5 remain better choices for complex agentic coding tasks, and that Haiku 5.5 is best suited to narrowly scoped tasks that might otherwise have been cost-prohibitive. It is the first Haiku-class model with an adjustable effort setting, according to the post.
OSWorld 2.1 (offline subset) scores from Anthropic’s table of 7 October 2026: Haiku 4.5 15.7%, GPT-6 Luna 48.9%, Haiku 5.5 72.4%, Sonnet 5.5 83.9%. Anthropic’s own evaluation.
The Credits and the Platform
Anthropic says it will roll out a monthly API credit this week to Max and Team subscribers: $100 a month for Max 5x users, $200 for Max 20x users and up to $500 for Team subscribers, pooled across their users. The credits can be used on any of its models.
Haiku 5.5 is available on all platforms, including Amazon Web Services, Google Cloud and Microsoft Azure, the post says. Anthropic is also adding computer use and browser use support, in beta, to its Python and TypeScript SDKs.
On safety, Anthropic says Haiku 5.5’s cybersecurity safeguards are more restrictive than Haiku 4.5’s but somewhat less restrictive than those on its other recent models, and still block penetration testing.
The Structural Read
The unit price falls further than the average. On prompts up to 100,000 tokens, input and output cost a tenth of Haiku 4.5’s on our arithmetic, while Anthropic’s own average for the cost to run is around 75% less.
Long prompts are priced differently. Above 100,000 tokens, Haiku 5.5 charges $0.50 for input and $2.50 for output, five times its own short-prompt rate.
The cache-read cut is aimed at agents. Anthropic says cache reads make up a large share of models’ token consumption, which is why halving them takes around 20% off Sonnet 5.5 on most agentic work.
Anthropic, 7 October 2026
“Haiku 5.5 is available at a much lower price than Haiku 4.5. On average, it now costs around 75% less to run.”
Three Implications
A TENTH OF THE UNIT PRICE Haiku 5.5 lists input at $0.10 and output at $0.50 per million tokens on prompts up to 100,000.
CHEAPER AGENT LOOPS Sonnet 5.5 cache reads fall to $0.10 per million tokens from $0.20, Anthropic says.
CREDITS FOR BUILDERS Max and Team subscribers get $100 to $500 a month in API credits, according to the post.
The framework’s starting point: “The cost of intelligence is collapsing while the cost of deploying intelligence is exploding.”
As we read it, Haiku 5.5 is the first half of that sentence in a price list, a tenth of the unit price on short prompts. Anthropic’s own advice to use it for subagents, compaction and summaries is the second half: cheaper tokens are meant to be used in far greater numbers.
What Is Not Established
We read Anthropic’s announcement; the footnotes to its speed and cost claims did not come through in the version we read, and we did not read the system card. The benchmark figures are Anthropic’s own, and we did not test the model or contact Anthropic.
Whether a given workload costs 75% less depends on its mix of prompt length, cache use and output; the post gives unit prices and an average, not totals for any customer.
Business Engineer Framework
The CFO’s Guide to the Token Economy
A Business Engineer framework on why the cost of intelligence falls while the cost of deploying it rises.
Anthropic priced Claude Haiku 5.5 at $0.10 per million input tokens and $0.50 per million output tokens on prompts up to 100,000 tokens, says it costs around 75% less to run than Haiku 4.5 on average, and halved Sonnet 5.5’s cache-read price to $0.10. It also added monthly API credits of $100 to $500 for Max and Team subscribers.
A note on sourcing. We read Anthropic’s Claude Haiku 5.5 announcement of 7 October 2026; its footnote text did not come through and we did not read the system card. Prices and benchmark scores are Anthropic’s; the 90% figure is our arithmetic. We did not test the model or contact Anthropic. Nothing here is a forecast, and nothing here is financial or investment advice.
Gennaro is the creator of FourWeekMBA, which reached about four million business people, comprising C-level executives, investors, analysts, product managers, and aspiring digital entrepreneurs in 2022 alone | He is also Director of Sales for a high-tech scaleup in the AI Industry | In 2012, Gennaro earned an International MBA with emphasis on Corporate Finance and Business Strategy.
Scroll to Top
Discover more from FourWeekMBA
Subscribe now to keep reading and get access to the full archive.
