Claude Haiku 4.5
Fast / cheapLast verified: September 25, 2026· Anthropic Claude Haiku 4.5 pricing
Anthropic’s fastest and most cost-effective model; 200K-token context window.
- Input / 1M
- $1.00
- Output / 1M
- $5.00
- Context
- 200K tokens
- Provider
- Anthropic
Pricing verified September 25, 2026. Prices change frequently. Always confirm against the provider’s official pricing page before relying on these figures for budgeting. Official pricing →
When to choose Claude Haiku 4.5
The right home for high-volume, latency-sensitive, low-stakes work: classification, extraction, routing, first-pass drafts. At $1 in and $5 out it is the cheapest Anthropic row, with a 200K window rather than the 1M the larger models carry — ample for the tasks above, short of what whole-repository reasoning needs. The saving is only real if evals show the answers held, which is why the boundary between this tier and the next belongs in code rather than in a habit.
Claude Haiku 4.5 sits close enough to Claude Sonnet 4.6 on price that the two are usually decided together. That comparison is written once, on the models table, rather than argued twice from opposite sides.
What Claude Haiku 4.5 costs in practice
At $1 and $5 the 50M/10M month comes to $100, and a request that fills the entire 200K window costs twenty cents of input. The figure worth noticing is that OpenAI o4-mini does the same month for $99 with the same 200K window, so those two are effectively interchangeable on price and the choice comes down to what each is actually good at. Below both, GPT-4o mini does it for $13.50, which is the reminder that a volume tier only pays when the work genuinely belongs there.
| Scenario | Input / mo | Output / mo | Est. cost / mo |
|---|---|---|---|
| Prototype | 2M | 0.5M | $5 |
| Growing product | 50M | 10M | $100 |
| At scale | 500M | 100M | $1,000 |
Plug in your own numbers with the cost calculator.
Where Claude Haiku 4.5 fits in a routing setup
As the cheapest Anthropic row this is normally the bottom of a Claude-only routing setup, but it is not the bottom of the table. GPT-4o mini does the reference month for $13.50 against $100, and DeepSeek-V3 for $24.50, so a genuinely price-driven route has further to fall if you are willing to cross vendors and, in DeepSeek’s case, accept an offshore API. Staying inside one vendor is a legitimate operational preference. It is worth knowing what it costs.
Cheaper alternatives to Claude Haiku 4.5
The models immediately below Claude Haiku 4.5 on blended cost, nearest first. A swap only pays if quality holds on the routes you move — see the LLM cost optimization playbook for how to prove that before you move them.
Frequently asked questions
How much does Claude Haiku 4.5 cost?
Claude Haiku 4.5 bills $1.00 per million input tokens and $5.00 per million output tokens — $2.00 per million blended at a three-to-one input:output split. A workload of 50M input and 10M output tokens per month works out at about $100.
What is Claude Haiku 4.5's context window?
Claude Haiku 4.5 has a 200K-token context window, of which up to 64K tokens can be the reply. One request that fills the whole window costs about $0.20 at the input rate, before the reply is billed.
Where does Claude Haiku 4.5 sit on price?
At $2.00 per million blended — three input tokens for every output token — Claude Haiku 4.5 is dearer than 7 of the 19 models tracked here and cheaper than 11. Output bills at 5x the input rate, and the nearest cheaper model on this table is OpenAI o4-mini at $1.93 blended.
Related concepts: context window, model routing, prompt caching, evals.