Curated models for your best work
Join the waitlistClaude Haiku 4.5
- Vision
- Tool calling
- Reasoning
- Streaming
- 2 providers
- Context window
- 200K
- tokens
- Input price
- $1.00
- per 1M tokens
- Output price
- $5.00
- per 1M tokens
- Max output
- 64K
- tokens
Provider list price
Pricing
USD / 1M tokens
| Input | $1.00 / M tokens |
| Output | $5.00 / M tokens |
| Cache read | $0.1 / M tokens |
| Cache write (5 min) | $1.25 / M tokens |
| Cache write (1 hour) | $2 / M tokens |
| Context window | 200K tokens |
| Routing | Free |
Provider list price. BYOK bills your provider; managed credits draw down at these rates. No markup or request fee. Every request is metered per engineer and joined to shipped work.
Registry 2026-10-31.2 · effective 9/17/2026 · Anthropic API pricing
Providers
Claude Haiku 4.5 runs on 2 providers. Automatic uses the cheapest one that is working, or you can choose. The list price applies unless a provider shows its own price.
No provider needed. TokenSpend picks the cheapest working one.
curl https://api.tokenspend.dev/v1/chat/completions \
-H "Authorization: Bearer $TOKENSPEND_API_KEY" \
-H "content-type: application/json" \
-d '{"model":"anthropic/claude-haiku-4-5","messages":[{"role":"user","content":"hello"}]}'Tool calls and JSON schema only go to providers that have passed our tests for them. Only Anthropic has passed for tools so far, so tool requests have no fallback.
Registry contract
Capabilities
- Protocols
- 3
- Chat · Responses · Messages · streaming
- Tools
- 2
- Chat · Messages
- JSON Schema
- None
- Not supported
- Vision
- Yes
- Chat · Responses · Messages
- Reasoning
- 1
- Messages
- Output limit
- 64,000
- tokens
canonical anthropic/claude-haiku-4-5 · revision anthropic/claude-haiku-4-5 · aliases claude-haiku-4-5
First request
Run it
from tokenspend import TokenSpend
client = TokenSpend()
resp = client.chat.completions.create(
model="anthropic/claude-haiku-4-5",
messages=[{"role": "user", "content": "hello"}],
)import { TokenSpend } from "@tokenspend/sdk";
const client = new TokenSpend();
const resp = await client.chat.completions.create({
model: "anthropic/claude-haiku-4-5",
messages: [{ role: "user", content: "hello" }]
});curl https://api.tokenspend.dev/v1/chat/completions \
-H "content-type: application/json" \
-H "Authorization: Bearer $TOKENSPEND_API_KEY" \
-d '{"model":"anthropic/claude-haiku-4-5","messages":[{"role":"user","content":"hello"}]}'export ANTHROPIC_BASE_URL=https://api.tokenspend.dev export ANTHROPIC_API_KEY=$TOKENSPEND_API_KEY export ANTHROPIC_MODEL=anthropic/claude-haiku-4-5
One request shape reaches every model. Claude Code connects as an integration with two environment variables. TokenSpend never stores prompts or responses.
Questions
Which provider will run my Claude Haiku 4.5 requests?
Anthropic and AWS Bedrock can all run it. By default TokenSpend sends each request to the cheapest one that is working and moves to the next if that one errors, so you do not have to pick. To pin a provider, send it in the request as the provider option; the code on this page shows how. Where a provider charges a rate of its own, you are billed that rate and it is shown above.
What happens if a provider goes down mid-request?
A failed attempt, such as a timeout, a rate limit or a server error, is retried once on the next provider before any output reaches you. A stream that has already started is not restarted on another provider. If you pinned one provider, requests stay on it and fail with a clear error instead of moving.
What does Claude Haiku 4.5 cost?
$1.00 per 1M input tokens and $5.00 per 1M output tokens. TokenSpend adds no markup and no per-request fee. If you pay with managed credits, buying them carries a 5% fee.
What can I do with Claude Haiku 4.5 through TokenSpend?
Chat and streaming, plus tool calling (Chat, Messages), image input (Chat, Messages) and reasoning controls (Messages). It does not support JSON schema output here. Tool calling and JSON schema are enabled per provider, only after that provider passes our tests, so a provider chosen by hand may lack them.
How long can my prompt be?
200K tokens of context, prompt and reply together, with up to 64,000 tokens in a single reply.
Does prompt caching make Claude Haiku 4.5 cheaper?
Yes, when the provider caches a repeated prompt prefix. Cached input is billed at $0.10 per 1M tokens instead of $1.00, and writing to the cache costs $1.25 per 1M tokens.
How current is this pricing?
It follows rate card 2026-10-31.2, effective 9/17/2026. Source: Anthropic API pricing.