The short version
Haiku 5.5 is built for work where volume and speed matter more than frontier reasoning. Anthropic says it reliably handles "quick and repetitive workloads" — summaries, compactions, database queries, and classification requests — and pairs well with Opus 5.5 and Sonnet 5.5 as a subagent on coding work. Because it's also the company's fastest model to date, it's aimed at speed-sensitive tasks like live customer support and browser use[1].
The biggest story is price. Haiku 5.5 costs around 75% less to run than Haiku 4.5 on average — for requests up to 100,000 tokens it's priced 90% lower, and for larger requests 50% lower[1].
Performance: a leap over Haiku 4.5
Haiku 5.5 marks a substantial step up from its predecessor across the board[1]:
| Benchmark | Haiku 5.5 | Haiku 4.5 |
|---|---|---|
| GDPval-AA v2.1 | 1620 | 735 |
| OSWorld 2.1 (offline) | 72.4% | 15.7% |
| Terminal-Bench 4.0 | 39.2% | 0.0% |
| Chartography (no tools) | 46.4% | 6.4% |
| Humanity's Last Exam (no tools) | 45.9% | 10.2% |
Anthropic-reported benchmark scores, October 2026[1].
Haiku 5.5 is also the first Haiku-class model with an adjustable effort setting, so developers can tune it toward cost or capability the same way they do with larger Claude models[1].
Pricing: from $0.10 per million input tokens
Anthropic introduced a two-tier price for Haiku 5.5 based on prompt size — the vast majority of requests fall into the cheaper tier[1]:
| Per 1M tokens | Haiku 5.5 (≤100k) | Haiku 5.5 (>100k) | Haiku 4.5 |
|---|---|---|---|
| Input | $0.10 | $0.50 | $1.00 |
| Output | $0.50 | $2.50 | $5.00 |
| Cache reads | $0.01 | $0.05 | $0.10 |
| Cache writes | $0.125 | $0.625 | $1.25 |
API pricing per 1M tokens, October 2026[1].
Claude Haiku 5.5 is available now on all platforms — including Amazon Web Services, Google Cloud, and Microsoft Azure — under the model ID claude-haiku-5-5[1].
What else changed
Anthropic bundled the launch with two value improvements for the wider model range[1]:
- Cheaper Sonnet 5.5 cache reads. Cache reads on Sonnet 5.5 dropped 50%, from $0.20 to $0.10 per million tokens — cutting the cost of most agentic Sonnet 5.5 work by roughly 20%.
- Monthly API credits. Claude Max and Team subscribers now get monthly API credits for the Claude Platform: $100 for Max 5x, $200 for Max 20x, and up to $500 pooled for Team users.
Bottom line
Verdict: If you're running high-volume, cost-sensitive Claude workloads — summaries, lookups, classification, subagents, or speed-critical support — Haiku 5.5 is an easy win over Haiku 4.5: dramatically cheaper, faster, and more capable. For complex agentic coding, Opus 5.5 and Sonnet 5.5 remain the better fit.
Frequently asked questions
When did Claude Haiku 5.5 release?
Anthropic announced Claude Haiku 5.5 on October 7, 2026[1][2].
How much does Claude Haiku 5.5 cost?
For prompts up to 100,000 tokens, $0.10 per million input tokens and $0.50 per million output tokens. Larger prompts cost $0.50 input and $2.50 output[1].
Is Haiku 5.5 cheaper than Haiku 4.5?
Yes — around 75% cheaper to run on average, and priced 90% lower for requests up to 100,000 tokens[1].
Where is Claude Haiku 5.5 available?
On all platforms, including AWS, Google Cloud, and Microsoft Azure, using the model ID claude-haiku-5-5[1].
Sources
- Anthropic — "Introducing Claude Haiku 5.5" — anthropic.com
- Claude Help Center — Release notes (October 2026) — support.claude.com
- Anthropic — "Claude Haiku 5.5 System Card" — anthropic.com