Home/Anthropic/Claude Haiku 5.5
Anthropic pricing

Claude Haiku 5.5 API Pricing & Cost Calculator

Claude Haiku 5.5 costs $0.10 input / $0.50 output per 1M tokens for prompts up to 100K tokens, and $0.50 / $2.50 above that — it is the one current Claude model without flat 1M-context pricing, so long prompts cost 5x more per token. Within 100K it is 90% cheaper per call than Claude Haiku 4.5 ($1.00/$5.00); Anthropic puts the average saving at about 75%. It uses the newer tokenizer, which counts roughly 30% more tokens for the same text, so measure your own prompts. Built for classification, extraction and routing; Sonnet 5.5 is the better choice for agentic coding. Cached input $0.01, Batch $0.05/$0.25. The calculator shows the up-to-100K price.

Input
$0.10
per 1M tokens
Output
$0.50
per 1M tokens
Context window
1M
tokens
Released
2026-10
≈ Estimated tokenizer·$0.10 in·$0.50 out (per 1M)
Quick start with a use case
Total cost per call$0.000350
Input$0.000100
Output$0.000250
Cost comparison
Standard
$0.000350
With Caching
$0.000305
Save 13% ↓
With Batch
$0.000175
Save 50% ↓
Cutting costs? Open models start at $0.28/1M

DeepSeek, Kimi & 200+ open models on Novita AI — often 5–20x cheaper than proprietary APIs for comparable tasks.

Try Novita AI

Affiliate link — we may earn a commission at no extra cost to you.

Detailed pricing

Claude Haiku 5.5 pricing breakdown

All pricing dimensions including caching and batch discounts.

TypePrice (per 1M tokens)
Input$0.1000
Output$0.5000
Cached input$0.0100
Cache write$0.1250
Batch input$0.0500
Batch output$0.2500
Cutting costs? Open models start at $0.28/1M

DeepSeek, Kimi & 200+ open models on Novita AI — often 5–20x cheaper than proprietary APIs for comparable tasks.

Try Novita AI

Affiliate link — we may earn a commission at no extra cost to you.

Last verified 2026-10-31 · Anthropic official pricing · ⚠️ Spotted a wrong price? Report in 30s →

How it compares

Claude Haiku 5.5 vs alternatives

Single-call cost (1000 input + 500 output tokens) ranked from cheapest.

ModelPer call
Claude Haiku 5.5
Anthropic · this page
$0.000350
GPT-6 Luna
OpenAI
$0.000350
DeepSeek V3.2
DeepSeek
$0.000480
GPT-5.6 Luna
OpenAI
$0.000800
DeepSeek V4-Flash
DeepSeek
$0.000900
Gemini 3.1 Flash Lite
Google
$0.001000
Recommended use

When to choose Claude Haiku 5.5

Claude Haiku 5.5 shines for general-purpose tasks, high-throughput, low-latency tasks, classification and routing, structured data extraction, and long documents and large codebases. Token counts are estimated within ~10-20% margin.

✓Context window of 1M tokens handles entire codebases or book-length documents.
✓Prompt caching available — significant savings for repeated system prompts.
✓Batch API support for non-realtime workloads at ~50% discount.
✓Tool / function calling supported.
FAQ

Frequently asked questions

Claude Haiku 5.5 costs $0.10 per 1M input tokens and $0.50 per 1M output tokens. A typical chat call (1000 input + 500 output tokens) costs approximately $0.0003. Use the calculator above to estimate your specific use case.
Claude Haiku 5.5 supports a 1M token context window with a max output of 128,000 tokens.
Yes. Cached input is priced at $0.01 per 1M tokens — 90% cheaper than uncached input. This is especially valuable for repeated system prompts, long-context retrieval, and chat threads with shared history.
Yes. Claude Haiku 5.5 supports the Batch API at $0.05 input / $0.25 output per 1M tokens — typically 50% off standard pricing. Batch jobs complete within ~24 hours, ideal for non-realtime workloads like overnight data processing or content generation pipelines.
Yes, Claude Haiku 5.5 accepts image input. Vision token pricing is generally calculated based on image dimensions and folded into the input token count. Specific image-specific pricing varies — refer to Anthropic's official documentation.
Token counts for Claude Haiku 5.5 are estimated from character ratios (~10-20% margin). Cost calculations use prices verified on 2026-10-31. For final billing accuracy, always verify with Anthropic's usage dashboard.
Claude Haiku 5.5 does not currently support fine-tuning. Consider another model in the Anthropic family if fine-tuning is required.
Related

Other models you might consider

OpenAI
GPT-6 Luna
The cheapest model on this site — with a 922K context window
$0.10 in$0.50 out
DeepSeek
DeepSeek V3.2
DeepSeek's open-weight MoE — top cost/quality ratio
$0.28 in$0.40 out
OpenAI
GPT-5.6 Luna
The cheapest long-context model OpenAI ships — 20x under the flagship
$0.20 in$1.20 out
DeepSeek
DeepSeek V4-Flash
1M context at budget pricing — $0.30/$1.20
$0.30 in$1.20 out
Google
Gemini 3.1 Flash Lite
Google's cheapest listed model — $0.25/1M input
$0.25 in$1.50 out