Skip to contentATP Token

Claude Haiku 4.5

AnthropicAll Anthropic models

Input

$1.00

USD / 1M tokens

Output

$5.00

USD / 1M tokens

Context

200K

tokens

Released

2025-10-15

Positioning: Fast. Price level 3 of 4 among text models. Vendor-stated speed: Fastest.

Claude Haiku 4.5 is Anthropic's small, low-latency model, released October 15, 2025, that its docs call the fastest Claude with near-frontier intelligence. Anthropic says it offers coding performance similar to Claude Sonnet 4 at one-third the cost and more than twice the speed, which suits real-time chat and sub-agent work.

ToolsVision
Get started freeAll model pricesCompare

Anthropic's Claude Haiku 4.5 model page lists a 200K-token context window and 64K max output tokens.

Pricing

Pay for what you use: credits are deducted only when a request succeeds.
Prices in USD · $1 = 100 credits
Input
$1.00
Cache read
$0.10
Output
$5.00
USD / 1M tokens
Cache write · 5m TTL$1.25 / 1M
Cache write · 1h TTL$2.00 / 1M

Anthropic · Family · USD / 1M tokens

Text models from the same vendor on ATP. Bars fill up to the family's highest price.
All Anthropic models
ModelInputOutput
Claude Fable 5claude-fable-5$10.00$50.00
Claude Opus 4.7claude-opus-4-7$5.00$25.00
Claude Opus 4.8claude-opus-4-8$5.00$25.00
Claude Opus 5claude-opus-5$5.00$25.00
Claude Sonnet 4.6claude-sonnet-4-6$3.00$15.00
Claude Sonnet 5claude-sonnet-5$3.00$15.00
Claude Haiku 4.5 This pageclaude-haiku-4-5$1.00$5.00

Cost per 1,000 typical requests

A quick estimate with fixed token counts per request. No caching, no retries. Try the estimator with your own numbers.
WorkloadUSD / 1,000
Short chat turn800 / 400$2.80
RAG answer4,000 / 500$6.50
Agent step with tools20,000 / 1,500$27.50

Where it fits

Good fit
  • Real-time chat assistants, customer-service agents, and in-app helpers
  • Classification, extraction, and routing at high volume
  • Parallel sub-agents working under a Sonnet- or Opus-class orchestrator
Usually not ideal
  • Deep multi-step reasoning or hard debugging (step up to Claude Sonnet 5)
  • Single requests that need more than a 200K-token window

Benchmarks

Vendor-reported
  • SWE-bench Verified73.3%

    bash + edit tools; averaged over 50 trials, 128K thinking budget, full 500-problem set, prompt addition

    Anthropic
  • Terminal-Bench41.0%

    Terminus 2; avg of 11 runs (6 without thinking, 5 with 32K thinking budget)

    Anthropic
  • OSWorld50.7%

    official OSWorld-Verified framework, 100 max steps, avg of 4 runs, 128K total thinking budget

    Anthropic
  • GPQA Diamond73.0%

    avg of 10 runs, 128K thinking budget

    Anthropic

Scores as published by each vendor, not measured by ATP. Test conditions vary, so they are not directly comparable across vendors.

Same controls on every model: project keys, per-project model access, credit budgets, and a log of every request.Set up your organization

Cost estimator

Plug in your usage to see roughly what a month costs.
0200
0100
ESTIMATE
$20.00
USD / month
2,000 credits / month
  • Input$10.00
  • Output$10.00

$20.00 USD / month · 2,000 credits / month

An estimate at our USD prices (USD 1 = 100 credits). What each request actually cost shows up in your console's request log. Cached input is billed at the lower cache-read price, so you may spend less.

Quickstart

Keep your SDK. Point it at ATP and use a project key. Read the docs
quickstart.py · OpenAI SDK
from openai import OpenAI

client = OpenAI(
  base_url="https://api.atptoken.ai/v1",
  api_key="atp-•••"
)

r = client.chat.completions.create(
  model="claude-haiku-4-5",
  messages=[{"role":"user","content":"hi"}],
)

Compare with similar models

Same type and positioning, from other vendors, closest in price first.
Compare side by side

Related models

All Anthropic models
Claude Fable 5claude-fable-5$10.00 / $50.00In / out · per 1M tokensClaude Opus 5claude-opus-5$5.00 / $25.00In / out · per 1M tokensClaude Opus 4.8claude-opus-4-8$5.00 / $25.00In / out · per 1M tokensClaude Opus 4.7claude-opus-4-7$5.00 / $25.00In / out · per 1M tokensClaude Sonnet 5claude-sonnet-5$3.00 / $15.00In / out · per 1M tokensClaude Sonnet 4.6claude-sonnet-4-6$3.00 / $15.00In / out · per 1M tokens

FAQ

4
Can Claude Haiku 4.5 use tools, read images, or think before answering?
Claude Haiku 4.5 supports tool calling and vision input, but not extended thinking or audio.
How much context does Claude Haiku 4.5 take?
Anthropic lists a 200,000-token context window for Claude Haiku 4.5. Source: Anthropic — Introducing Claude Haiku 4.5. The model list your key can see shows the limit you actually get.
Is cached input cheaper on Claude Haiku 4.5?
You pay per 1M input tokens and per 1M output tokens. Cached input costs 10% of the normal input price. Cache writes are priced separately. You only pay when a request succeeds. $1 = 100 credits. See pricing
Is there a newer Anthropic model than Claude Haiku 4.5 on ATP?
Yes. Claude Opus 5 (claude-opus-5) came out on Jul 24, 2026, after Claude Haiku 4.5 (Oct 15, 2025), and is the newer Anthropic text option on ATP. Both stay available: compare family prices on this page, then turn on the id you want for your project.

Sources

The model description on this page cites these sources. Prices come from ATP.
Get started

Start with $5. Credits never expire.

Create a project key, turn this model on, and pay only for what you use.

Get started freeTalk to enterprise