Announced September 28, 2026 alongside Sonnet 5.5; generally available October 7, 2026. Gateway id anthropic/claude-haiku-5.5. Tiered pricing: $0.10/$0.50 per 1M up to 100K input tokens, $0.50/$2.50 above. 1,000,000 context, 128,000 max output. Launch figures as reported by Anthropic: Terminal-Bench 4.0 39.2 (max effort), 31.6 (extra), 12.8 (low); OSWorld 2.1 offline subset 72.4 (Haiku 4.5: 15.7). Verified October 10, 2026.
$0.10/$0.50 per 1M below 100K input tokens — a twentieth of Sonnet 5.5
Effort is adjustable from none to max, so one model covers both cheap lookups and harder steps
Anthropic reports OSWorld 2.1 (offline subset) rising from 15.7% on Haiku 4.5 to 72.4%
Weaknesses
Long prompts cost five times more: above 100K input tokens the rate becomes $0.50/$2.50
Effort matters a great deal: Terminal-Bench 4.0 is 12.8% at low effort against 39.2% at max
Real-world use cases
What people actually use Claude Haiku 5.5 for.
Sub-agents in a larger agent system, where dozens of cheap calls replace one expensive one
Classification, extraction and routing at volume with thinking turned off
Light coding help — Anthropic reports 39.2% on Terminal-Bench 4.0 at max effort
How Claude Haiku 5.5 compares
The nearest models people weigh against it, and what actually separates them.
vs Claude Opus 5.5 — Against Claude Opus 5.5 (Anthropic), Claude Haiku 5.5 runs about 98% cheaper per token and answers faster. Take Claude Haiku 5.5 unless you specifically need what Claude Opus 5.5 does better.
vs Claude Sonnet 5.5 — Against Claude Sonnet 5.5 (Anthropic), Claude Haiku 5.5 runs about 95% cheaper per token and answers faster. Take Claude Haiku 5.5 unless you specifically need what Claude Sonnet 5.5 does better.
vs GPT-6 Luna — Against GPT-6 Luna (OpenAI), Claude Haiku 5.5 lands within a few percent on price and gives up 1.1x on context. Which one wins depends on whether context depth or latency is your constraint.
Ready to try it?
Start using Claude Haiku 5.5
High-volume tool use, sub-agents and everyday tasks on a tight budget. Start free — no card required.
Anthropic's September 22, 2026 Opus — a step up from Opus 5 on agentic coding, long-running agent work and vision, at a lower $4/$20 list price and with fewer tokens spent per finished task.
Verdict
Anthropic's new coding default — top SWE-bench Pro score at a lower price than Opus 5.
Quality score
98%
Pricing
$4.00/1M in
$20.00/1M out
Speed
Deliberate
2/5 speed
Context
1M tokens
Released September 22, 2026. Gateway id anthropic/claude-opus-5.5. $4/$20 per 1M (cache read $0.20, cache write $5); Fast mode $8/$40. 1,000,000 context, 128,000 max output, knowledge cutoff June 2026, effort levels low to max. Launch figures as reported by Anthropic: SWE-bench Pro 89.9 (Fable 5.1 81.2, Opus 5 79.2); SWE-bench Multilingual 93.9; Terminal-Bench 4.0 66.4 (xhigh); OSWorld 2.1 81.8; Humanity's Last Exam 67.7; Toolathlon Verified 77.8 (Opus 5 80.6). Verified October 10, 2026.
CodingAgenticFlagship1M contextPremiumNew
Best for
Agentic coding, long-running agents and knowledge work where Opus 5 was the bar
Anthropic's September 28, 2026 Sonnet — built for well-scoped everyday work like features, bug fixes, documents, slides and spreadsheets, priced at $2/$10 and close to Opus 5.5 on most of Anthropic's launch rows.
Verdict
Near-Opus 5.5 quality on scoped work at half the price.
Quality score
96%
Pricing
$2.00/1M in
$10.00/1M out
Speed
Fast
4/5 speed
Context
1M tokens
Released September 28, 2026. Gateway id anthropic/claude-sonnet-5.5. $2/$10 per 1M (cache read $0.10, cache write $2.50). 1,000,000 context, 128,000 max output, knowledge cutoff June 2026. Launch figures as reported by Anthropic (Sonnet 5.5 / Opus 5.5): Terminal-Bench 4.0 70.6 / 66.4; OSWorld 2.1 80.1 / 81.8; Humanity's Last Exam 64.5 / 67.7; GDPval-AA v2.1 1,844 / 1,846. Verified October 10, 2026.
CodingWritingFast1M contextValueNew
Best for
Everyday feature work, bug fixing and polished documents at mid-tier pricing
OpenAI's September 22, 2026 low-cost reasoning model, replacing GPT-5.6 Luna at $0.10/$0.50 per 1M — half the predecessor's price, with the same 1.05M context.
Verdict
OpenAI's budget tier at half the old price.
Quality score
80%
Pricing
$0.10/1M in
$0.50/1M out
Speed
Very fast
5/5 speed
Context
1.1M tokens
Released September 22, 2026 with GPT-6 Sol. Gateway id openai/gpt-6-luna. $0.10/$0.50 per 1M up to 272K input tokens, $0.20/$0.75 above; Flex $0.05/$0.25. 1,050,000 context, 128,000 max output, knowledge cutoff May 18, 2026. Verified October 10, 2026.
BudgetFastHigh volumeNew
Best for
High-volume, focused tasks where cost per call decides the model
Pricing moves, ranking shifts, and capability updates.
New ModelOct 7, 2026
Claude Haiku 5.5 generally available — adjustable reasoning at $0.10/$0.50
Anthropic made Claude Haiku 5.5 generally available on October 7, 2026, after announcing it with Sonnet 5.5 on September 28. It is the first Haiku with adjustable reasoning effort, from off to max. Pricing is tiered: $0.10 per million input tokens and $0.50 per million output up to 100K input tokens, and $0.50/$2.50 above that. It has a 1M-token context and 128K output. Anthropic reports 39.2% on Terminal-Bench 4.0 at max effort (12.8% at low effort) and 72.4% on the offline subset of OSWorld 2.1, against 15.7% for Haiku 4.5. Artificial Analysis scores it 43 on its Intelligence Index, ahead of Gemini 3.8 Flash (41) and GPT-6 Luna (38). Verified October 10, 2026.
Claude Haiku 5.5 costs $0.1 per million input tokens and $0.5 per million output tokens on the API, with cached input at $0.01 per million. A month of 10M input and 2M output tokens runs about $2.00 at list price, before any batch or caching discounts.
What is the context window of Claude Haiku 5.5?
Claude Haiku 5.5 has a 1M tokens context window, with up to 128k tokens of output per response. That is the total of prompt plus response the model can hold in one request.
What is Claude Haiku 5.5 best for?
Claude Haiku 5.5 is best for high-volume tool use, sub-agents and everyday tasks on a tight budget. It is a strong fit when that workflow matters more than the tradeoffs around budget pricing and very fast speed.
When should I avoid Claude Haiku 5.5?
Your prompts routinely exceed 100K tokens, where the price quintuples, or the task needs sustained reasoning that Sonnet 5.5 handles far better.
What is a cheaper alternative to Claude Haiku 5.5?
Claude Sonnet 5.5 (Anthropic) at $2.00/1M/1M input against Claude Haiku 5.5's $0.10/1M/1M. Near-Opus 5.5 quality on scoped work at half the price. Compare it first if Claude Haiku 5.5's pricing is the thing stopping you.
What is a faster alternative to Claude Haiku 5.5?
Claude Opus 5.5 — deliberate against Claude Haiku 5.5's very fast, with 1M tokens of context. Worth the swap when response time is what your users notice rather than the last few points of reasoning depth.
Newsletter
Get notified when Claude Haiku 5.5 pricing changes
We track pricing daily. When this model drops or spikes, you'll know first.
No spam. Useful updates only. Affiliate disclosures always clearly labeled.