A8gent
GlossaryAgent concepts

What is claude Haiku 4.5?

Also called: claude-haiku-4-5, claude haiku 4.5 pricing, haiku 4.5 api price, cheapest claude model

Claude Haiku 4.5 is Anthropic's fast, cheapest tier, with a 200,000-token context window and a published API price of $1.00 per million input tokens and $5.00 per million output tokens.

Written by A8gent
Contents
  1. At a glance
  2. What Claude Haiku 4.5 is
  3. What it costs
  4. Where the cheap tier is the right answer
  5. Where it is the wrong answer
  6. What you should know
  7. FAQ
  8. Related terms

At a glance

Model ID
claude-haiku-4-5
Input price
$1.00 per 1M tokens (rate table dated June 2026)
Output price
$5.00 per 1M tokens
Context window
200,000 tokens
Positioned as
Fast, low-cost tier for high-volume routine work

What Claude Haiku 4.5 is

claude-haiku-4-5 is the model ID for Anthropic's fast, low-cost tier. It is the cheapest Claude model you can call, and for a large share of real automation work it is also the correct one.

Its context window is 200,000 tokens, which is materially smaller than the 1,000,000 carried by claude-sonnet-5 and claude-opus-5. That is the real constraint to check before choosing it: not capability on your task, but whether your prompt fits.

What it costs

On Anthropic's published rate table dated June 2026, claude-haiku-4-5 is $1.00 per million input tokens and $5.00 per million output tokens.

That is a fifth of Opus 5 on input and output alike, and a third of Sonnet 5's standard rate. Same support-agent shape as the other entries here, 5,700 input and 400 output tokens, works out at about 0.77 cents a turn, or roughly $460 a month across ten thousand six-turn conversations.

Where the cheap tier is the right answer

Classification, routing, extraction, tagging, short structured outputs and filling a schema are all jobs where the frontier tier rarely earns its multiple. The way to find out is not intuition, it is an eval set built from your own inputs: run Haiku and Sonnet side by side, score them, and only pay more where the score justifies it.

A common and effective pattern is routing. Send everything to Haiku first, and escalate to claude-sonnet-5 only on the inputs a cheap classifier flags as hard. Most teams find the majority of traffic never needs to escalate.

Where it is the wrong answer

Long-horizon agentic work, hard multi-step reasoning, and anything where a quiet wrong answer is expensive downstream. Also anything that simply will not fit: at 200,000 tokens the window rules out whole-corpus prompts that the 1M-token tiers can take, so check the prompt size before the price.

And as always the token bill is only part of it. Every agent step is also one n8n execution or one Zapier task, billed separately from tokens.

What you should know

These figures come from Anthropic's published rate table dated June 2026, not from a live reading of the pricing page, which we could not reach when this entry was written. Treat claude.com/pricing as the authority.

Claude Haiku 4.5 FAQ

How much does Claude Haiku 4.5 cost?

Anthropic's published rate table dated June 2026 lists $1.00 per million input tokens and $5.00 per million output tokens, making it the cheapest Claude model available.

What is the context window of Claude Haiku 4.5?

200,000 tokens. That is smaller than the 1,000,000-token window on Claude Sonnet 5 and Claude Opus 5, so check that your prompt fits before choosing it on price alone.

Is Claude Haiku 4.5 good enough for production?

For classification, routing, extraction and short structured outputs it usually scores the same as more expensive tiers on a real eval set. Build that eval set from your own inputs rather than assuming either way.

What is the cheapest Claude model?

Claude Haiku 4.5, at $1.00 per million input and $5.00 per million output tokens on the June 2026 rate table. It is a fifth of Claude Opus 5 on both input and output.

Primary source

Anthropic pricing

Checked June 2026. Anything priced or rate-limited moves, so verify against the source before you budget.

Want this built rather than explained?

Knowing what claude haiku 4.5 means is one thing; having it running in your business is another. We scope one workflow at a time, quote a fixed price in writing after a free consult, build it, and hand it over with documentation you own.