Claude Haiku 5.5 API

POST Base URL: https://api.hiapi.ai /v1/chat/completions

This model uses an OpenAI-compatible Chat Completions endpoint. The same HiAPI API key can call enabled models in your account group; image, video, and audio models use /v1/tasks and a different request shape.

Model overview

Model name	Claude Haiku 5.5
Provider	Anthropic
Context / max output	1M / 128K tokens
Endpoint	POST /v1/chat/completions
Reasoning	low / medium / high / xhigh · default medium
Pricing	$0.15 / $0.75 input/output; cached read $0.015; cache write $0.1875 (5m) or $0.30 (1h) per 1M tokens up to 100K input

Call Anthropic's Claude Haiku 5.5 through HiAPI's OpenAI-compatible Chat Completions API for fast text and coding workflows.

Production guidance

Request contract

Use model=claude-haiku-5-5 with POST /v1/chat/completions.
Use reasoning_effort=low, medium, high, or xhigh; medium is the default.
In the Playground, choose 5m or 1h only after entering a non-empty system prompt. Put cache_control={type:"ephemeral"} inside the system text content block and omit ttl for five minutes; set ttl:"1h" for one hour. Omit cache_control entirely when caching is off. Never put it at the top level.
This contract uses native OpenAI-compatible Chat Completions fields. Image input, tools, JSON mode, and the Responses API are not advertised until separately verified.

Best suited for

Fast agentic work

Text, coding, and short multi-step workflows where controllable reasoning effort and latency both matter.

messagesreasoning_effort

Request parameters

model string required

Use this exact public model ID.

example claude-haiku-5-5 enum: claude-haiku-5-5

messages array required

Ordered system, user, and assistant messages.

stream boolean optional

Set true for Server-Sent Events.

default false

reasoning_effort enum optional

Use low, medium, high, or xhigh in the native Chat Completions request.

default medium enum: lowmediumhighxhigh

max_tokens integer optional

Maximum visible output and thinking budget.

API examples

POST https://api.hiapi.ai/v1/chat/completions

# Call a text model; a non-streaming request returns one Chat Completions response
# Replace YOUR_API_KEY with your account API key
curl -X POST "https://api.hiapi.ai/v1/chat/completions" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "claude-haiku-5-5",
  "messages": [
    {
      "role": "system",
      "content": [
        {
          "type": "text",
          "text": "You are a concise migration assistant.",
          "cache_control": {
            "type": "ephemeral",
            "ttl": "5m"
          }
        }
      ]
    },
    {
      "role": "user",
      "content": "Summarize this migration plan in three risks and three mitigations."
    }
  ],
  "stream": false,
  "reasoning_effort": "medium",
  "max_tokens": 4096
}'

import requests, json

# OpenAI-compatible Chat Completions request body
payload = json.loads(r'''
{
  "model": "claude-haiku-5-5",
  "messages": [
    {
      "role": "system",
      "content": [
        {
          "type": "text",
          "text": "You are a concise migration assistant.",
          "cache_control": {
            "type": "ephemeral",
            "ttl": "5m"
          }
        }
      ]
    },
    {
      "role": "user",
      "content": "Summarize this migration plan in three risks and three mitigations."
    }
  ],
  "stream": false,
  "reasoning_effort": "medium",
  "max_tokens": 4096
}
''')

# Call a text model; a non-streaming request returns one Chat Completions response
resp = requests.post(
    "https://api.hiapi.ai/v1/chat/completions",
    headers={"Authorization": "Bearer YOUR_API_KEY", "Content-Type": "application/json"},
    json=payload,
)
print(resp.json())  # Read text from choices[0].message.content and Token usage from usage

// Call a text model; a non-streaming request returns one Chat Completions response
// Replace YOUR_API_KEY with your account API key
const resp = await fetch("https://api.hiapi.ai/v1/chat/completions", {
  method: "POST",
  headers: {
    "Authorization": "Bearer YOUR_API_KEY",
    "Content-Type": "application/json",
  },
  body: JSON.stringify({
  "model": "claude-haiku-5-5",
  "messages": [
    {
      "role": "system",
      "content": [
        {
          "type": "text",
          "text": "You are a concise migration assistant.",
          "cache_control": {
            "type": "ephemeral",
            "ttl": "5m"
          }
        }
      ]
    },
    {
      "role": "user",
      "content": "Summarize this migration plan in three risks and three mitigations."
    }
  ],
  "stream": false,
  "reasoning_effort": "medium",
  "max_tokens": 4096
}),
});
console.log(await resp.json()); // Read text from choices[0].message.content and Token usage from usage

Request examples

Streaming

Read content deltas until [DONE], then inspect usage when returned.

Request body

{
  "model": "claude-haiku-5-5",
  "messages": [
    {
      "role": "system",
      "content": [
        {
          "type": "text",
          "text": "You are a concise migration assistant.",
          "cache_control": {
            "type": "ephemeral",
            "ttl": "5m"
          }
        }
      ]
    },
    {
      "role": "user",
      "content": "Summarize this migration plan in three risks and three mitigations."
    }
  ],
  "stream": true,
  "reasoning_effort": "medium",
  "max_tokens": 4096
}

Response schema

Read visible text from choices[0].message.content. For streams, accumulate deltas until [DONE]. Usage is the billing source of truth.

{
  "id": "chatcmpl_example",
  "object": "chat.completion",
  "model": "claude-haiku-5-5",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "Start with a migration boundary that can be rolled back."
      },
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 32,
    "completion_tokens": 48,
    "total_tokens": 80
  }
}

Read choices[0].message.content for visible text.
For streams, accumulate content and reasoning deltas until [DONE].
Read input and output usage; thinking counts as output.

FAQ

What is Claude Haiku 5.5?

It is Anthropic's Claude Haiku 5.5. The verified model specification is a 1M-token context window and 128K maximum output; this page shows the Chat Completions request format.

Which endpoint and model ID should I use?

Use POST /v1/chat/completions with model=claude-haiku-5-5.

Which reasoning effort values are supported?

Use reasoning_effort=low, medium, high, or xhigh. Medium is the default. The field is part of the native Chat Completions request.

Does this integration support images, tools, JSON mode, or Responses?

The current public contract exposes text Chat Completions only. Those capabilities are not advertised for this model until separately verified.

How do I use prompt caching?

Enter a non-empty system prompt before choosing a cache duration in the Playground. Put cache_control inside the system text block: {"type":"ephemeral"} uses the default five-minute cache, while {"type":"ephemeral","ttl":"1h"} requests one hour. Omit cache_control for no explicit cache control; a top-level cache_control field is not supported.

How are tokens billed?

For input up to 100K tokens, rates per 1M tokens are $0.15 input, $0.75 output, $0.015 cached read, $0.1875 cache write for 5 minutes, and $0.30 for 1 hour. Above 100K, rates are $0.75, $3.75, $0.075, $0.9375, and $1.50. The API response reports token usage; HiAPI usage logs show the settled charge. View live pricing.