Models and prices

Claude Haiku 5.5

Anthropic · released October 7, 2026

anthropic/claude-haiku-5-5

Input

Cache read

90% below input

Output

Context

1M tokens

Max output

128K

without max_tokens: 16K

Reasoning

Fast · Balanced · Thorough · Intensive · Exhaustive

default balanced

Prompt cache

On cache_control marks in the request

Hosts

A call goes to the default host and is charged at the price of the host that answered. The provider parameter pins the host.

HostInput₽ per 1MCache read₽ per 1MOutput₽ per 1MFirst tokensThroughputtok/s
AnthropicdefaultProcesses data outside Russia

Cost example

Without cache and reasoning tokens: their number depends on the reasoning level.

How a call is billed

  • A call costs what the host that answered it charges, with no markup. Prices on this page are the hosts' own.
  • Prices in rubles at the Central Bank of Russia rate; a call is charged at the rate of the day of the call.
  • Before the call, an estimate by the dearest host of the route is reserved on the balance. After the answer, the actual cost is charged and the rest of the reserve is returned.
  • There is no top-up fee, and the unspent remainder is refunded in full.
Samreshuuu Router

Request example

Pass an organization key with the models:write scope in the Authorization header. The same address works with the OpenAI SDK and OpenRouter clients.

bash
curl https://api.samreshuuu.ru/v1/chat/completions \ -H "Authorization: Bearer $SAMRESHUUU_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"anthropic/claude-haiku-5-5","messages":[{"role":"user","content":"Hi! What can you do?"}]}'