Anthropic's Claude API models, the Anthropic LLM API on APIMart: Claude Fable 5.1, Opus 5.5, Sonnet 5.5 and Haiku 5.5, plus earlier Opus, Sonnet and Haiku versions, through the native Anthropic Messages API or an OpenAI-compatible endpoint with one API key.
Service Highlights
Guides, comparisons and pricing analysis from the APIMart blog.
What's on your mind?
Ask APIMart
claude-fable-5.2
Transparent pricing with no hidden fees. Pay only for what you use.
Per 1M tokens
| Rate | GoldCurrent price | Platinum | Diamond | Official | Your savings |
|---|---|---|---|---|---|
| Input | 64~$6.4 | 60.8~$6.08 | 57.6~$5.76 | 80~$8 | 20% |
| Cached input | 1.6~$0.16 | 1.52~$0.152 | 1.44~$0.144 | 2~$0.2 | 20% |
| Cache write (5 min) | 80~$8 | 76~$7.6 | 72~$7.2 | 100~$10 | 20% |
| Cache write (1 hr) | 128~$12.8 | 121.6~$12.16 | 115.2~$11.52 | 160~$16 | 20% |
| Output | 320~$32 | 304~$30.4 | 288~$28.8 | 400~$40 | 20% |
Call Anthropic's current Claude lineup on APIMart: Claude Opus 5.5 for most coding and agent work, Claude Fable 5.1 for the hardest reasoning, Claude Sonnet 5.5 for speed and intelligence, and Claude Haiku 5.5 for high-volume tasks. Use the native Messages API with the Anthropic SDK or the OpenAI-compatible endpoint, and pay per token.
What you get when you call Claude models through APIMart
How teams use Claude models in production
Three steps to your first Claude request on APIMart.
Create an APIMart account and top up your balance. It works for every model on the platform, including the Claude API.
Generate an API key in your dashboard. The same key calls every Claude model and all other models on APIMart.
Point the Anthropic SDK at https://api.apimart.ai and call /v1/messages with a model ID such as claude-opus-5-5, or try it in the playground above.
Send your first Claude Opus 5.5 request through APIMart using the native Messages API or an OpenAI-compatible client.
Create an APIMart API key, add funds and set the APIMART_API_KEY environment variable on your server. For Python, install anthropic or openai with pip. An Anthropic Console key cannot authenticate to APIMart.
Use https://api.apimart.ai with the Anthropic SDK, or https://api.apimart.ai/v1 with the OpenAI SDK. Both require an APIMart key and the exact model ID. Use Messages for Anthropic-specific request and response formats.
curl --fail-with-body https://api.apimart.ai/v1/messages \
-H "x-api-key: $APIMART_API_KEY" \
-H "anthropic-version: 2025-10-01" \
-H "Content-Type: application/json" \
-d '{
"model": "claude-opus-5-5",
"max_tokens": 1024,
"messages": [
{
"role": "user",
"content": "Explain when to use prompt caching in an AI assistant."
}
]
}'import os
from anthropic import Anthropic
client = Anthropic(
api_key=os.environ["APIMART_API_KEY"],
base_url="https://api.apimart.ai",
)
message = client.messages.create(
model="claude-opus-5-5",
max_tokens=1024,
messages=[{"role": "user", "content": "Explain when to use prompt caching in an AI assistant."}],
)
for block in message.content:
if block.type == "text":
print(block.text)import os
from openai import OpenAI
client = OpenAI(
api_key=os.environ["APIMART_API_KEY"],
base_url="https://api.apimart.ai/v1",
)
response = client.chat.completions.create(
model="claude-opus-5-5",
max_tokens=1024,
stream=False,
messages=[{"role": "user", "content": "Explain when to use prompt caching in an AI assistant."}],
)
print(response.choices[0].message.content)What developers say about our Claude API
“Claude's 200K context window lets us process entire contracts in a single API call. No more chunking.”
Alex Chen
Legal Tech Engineer
“We switched from GPT to Claude for customer support — the responses are more nuanced and less likely to hallucinate.”
Sarah Jones
Support Lead
“Claude Opus 4.6 is the best coding assistant I've used. It understands large codebases remarkably well.”
Mike Brown
Senior Developer
“The thinking models are fantastic for research tasks. They show their work, which builds trust with our team.”
Emily Zhang
Research Scientist
“Haiku gives us GPT-4 level quality at a fraction of the cost for our chatbot use case.”
David Li
Startup Founder
“APIMart's Claude API is more affordable and just as reliable as going direct. Easy migration.”
Lisa Wang
Platform Engineer
Choose between Claude Opus 5.5, Sonnet 5.5, Haiku 5.5 and Fable 5.1. Compare APIMart token rates, context windows and use cases, then select a model to open its playground and detailed pricing.
APIMart Gold member prices in USD per 1 million tokens. Input and output are billed separately; cache reads apply only to cache hits. Cache writes have separate rates in the detailed pricing above.
| Model / API ID | Input | Output | Cache read | Context window | When to choose it |
|---|---|---|---|---|---|
Claude Opus 5.5claude-opus-5-5 | $3.20 | $16.00 | $0.16 | 1,000,000Max output 128,000 | Complex coding, multi-step agents and knowledge work. |
Claude Sonnet 5.5claude-sonnet-5-5 | $1.60 | $8.00 | $0.16 | 1,000,000Max output 128,000 | Everyday coding and assistants with a lower token budget. |
Claude Haiku 5.5claude-haiku-5-5 | $0.08≤ 100,000 input tokens | $0.40 | $0.008 | 1,000,000Max output 128,000 | High-volume classification, extraction and routing. |
| $0.40> 100,000 input tokens | $2.00 | $0.04 | |||
Claude Fable 5.1claude-fable-5.1 | $8.00 | $40.00 | $0.20 | 1,000,000Max output 128,000 | Demanding reasoning and long-horizon agent tasks. |
APIMart routes Claude requests to Anthropic’s official API, so context windows, max output and model features match Anthropic’s published specs; only the price differs. A dash means the value is unavailable. For tiered models, the input token count determines the rates for the request.
Common questions about the Claude API on APIMart
APIMart offers Anthropic's current Claude models, Claude Fable 5.1, Claude Opus 5.5, Claude Sonnet 5.5 and Claude Haiku 5.5, plus Claude Opus 5, Sonnet 5, Fable 5, Opus 4.8, 4.7, 4.6 and 4.5, Sonnet 4.6 and 4.5, and Haiku 4.5. Use the model IDs listed in the table above, for example claude-opus-5-5, claude-sonnet-5-5 or claude-haiku-5-5. Note that APIMart's Fable 5.1 ID is claude-fable-5.1, with a dot.
Start with Claude Opus 5.5 for most coding, agent and knowledge work, as Anthropic recommends. Move up to Claude Fable 5.1 for the most demanding reasoning and long-horizon tasks, use Claude Sonnet 5.5 when you need faster responses at a lower price, and pick Claude Haiku 5.5 for high-volume classification, extraction and routing.
Claude Haiku 5.5, released October 7, 2026, is Anthropic's newest model. The current lineup also includes Claude Sonnet 5.5 (September 28, 2026), Claude Opus 5.5 (September 22, 2026) and Claude Fable 5.1 (September 1, 2026), and all four are available on APIMart.
Compare the Gold member prices in the table above, quoted in USD per million tokens. Each model has separate input, output and cache rates. For tiered models such as Haiku 5.5, the input token count determines the rate; cache writes are billed separately. Select a model for its full pricing and membership tiers.
The comparison table shows APIMart Gold member rates. The detailed pricing section shows the available membership tiers and reference prices. Compare the same model, token type and input-length tier; discounts and cache prices can differ by rate.
Match the model to the task: Claude Haiku 5.5 for classification and extraction, Sonnet 5.5 for everyday work, and Opus 5.5 or Fable 5.1 only where you need the extra capability. Use prompt caching for repeated system prompts and documents, keep Haiku 5.5 prompts under 100K tokens to stay in the lower tier, and upgrade your membership for up to 28% off.
No. APIMart does not offer free credits for the Claude API. Claude models are pay-as-you-go by token usage, you only pay for successful requests, and Gold members get 20% off all prices.
Sign up for APIMart, add funds, open the API Keys page in your dashboard and create a new key. The same key calls every Claude model and all other models on the platform, with pay-as-you-go billing and no subscription. An APIMart key is separate from a key created in Anthropic's own Claude Console.
Yes. Sign in to APIMart, create an API key and add funds, then use the playground at the top of this page to chat with Claude in your browser. Playground messages use your API key and are billed per token from your balance, at the same rates as the API.
No. Claude.ai is Anthropic's chat app with its own subscription plans, and a Claude.ai subscription does not include API usage. The Claude API is billed separately per token; on APIMart you use an APIMart API key and pay from your APIMart balance.
You can top up your APIMart balance with Visa or Mastercard, Google Pay, Apple Pay, Alipay, WeChat Pay or crypto. Billing is pay-as-you-go with no subscription and no top-up fee, and the same balance works for every Claude model.
Yes. Anthropic's API for its Claude models is called both the Claude API and the Anthropic API. On APIMart you call the same Anthropic models, such as Claude Opus 5.5 and Sonnet 5.5, with the same Messages API request format, but you sign up and pay through APIMart instead of an Anthropic account. Requests are processed by Anthropic’s official API, so model behavior and features are the same as calling Anthropic directly; only the price differs.
No. Requests to api.apimart.ai need an APIMart API key, and an Anthropic API key from the Claude Console only works with Anthropic's own endpoint. Create a free APIMart account, add funds and generate a key in a few minutes; it works for every Claude model and all other models on APIMart.
Yes. APIMart supports the native Claude Messages protocol at POST /v1/messages. Keep your Anthropic SDK code, set the base URL to https://api.apimart.ai, use your APIMart key as the API key and call a model ID such as claude-opus-5-5. Responses are passed through in Anthropic's format.
Use https://api.apimart.ai with the Anthropic SDK, or https://api.apimart.ai/v1 with the OpenAI SDK. Both require an APIMart key and the exact model ID. Use Messages for Anthropic-specific request and response formats.
Yes. Add cache_control to a stable prefix such as a system prompt, document or codebase. The default cache lasts 5 minutes; for a 1-hour cache, set ttl to 1h and send the anthropic-beta: extended-cache-ttl-2025-04-11 header. Caching works through both the Messages API and Chat Completions, and the usage fields show cache writes and cache hits.
Anthropic's current models use adaptive thinking, where the model decides how much to reason. On APIMart, prefer the standard model ID with the thinking parameter over the platform's -thinking aliases, which are kept for earlier models. Thinking tokens are included in output tokens, and in multi-turn conversations you must pass thinking blocks back with their signature unchanged.
Claude Opus 5.5 is Anthropic's recommended default for long-running agentic coding and knowledge work. Claude Fable 5.1 is slower and aimed at the most demanding reasoning and long-horizon agentic work, at a higher price: Anthropic lists Fable 5.1 at $10 / $50 and Opus 5.5 at $4 / $20 per million input and output tokens.
Opus 5.5 and Sonnet 5.5 are Anthropic's newer versions, and Anthropic publishes a migration guide for moving from Opus 5 and earlier. Opus 5.5 is also cheaper: Anthropic lists it at $4 / $20 per million input and output tokens versus $5 / $25 for Opus 5, while Sonnet 5.5 keeps Sonnet 5's $2 / $10 list price. Opus 5 and Sonnet 5 remain available on APIMart.
Claude Fable 5.1, Opus 5.5, Sonnet 5.5 and Haiku 5.5 have a 1M-token context window and up to 128K output tokens, as listed by Anthropic. APIMart routes Claude requests to Anthropic’s official API, so the same limits apply on APIMart.
Yes. Set ANTHROPIC_BASE_URL to https://api.apimart.ai, use your APIMart key as the auth token and choose a Claude model ID. The APIMart docs include a step-by-step Claude Code setup guide.
No. Under APIMart's Terms of Service, your data is not used to train AI models or shared with third parties except as needed to provide the service. Requests are processed only to serve your API calls.
Explore more models in the same category.
Gemini 4 Argon
Gemini-4-Argon is an AI assistant that brings together analytical thinking and creative expression. From untangling complex problems to refining writing and code, it offers clear reasoning, natural conversation, and practical suggestions to help turn rough ideas into actionable next steps.
Claude Haiku 5.5
Claude Haiku 5.5 API on APIMart delivers fast, cost-efficient AI, offering responsive conversations, capable coding assistance, and reliable performance for high-volume workflows.
Claude Opus 5
Claude Opus 5 is Anthropic’s next-generation flagship large language model, featuring enhanced capabilities in code development, complex reasoning, knowledge analysis, and agent execution, making it ideal for large-scale project development and professional work scenarios.
Kimi K3
Kimi-K3 is a next-generation large language model launched by Moonshot AI. It features ultra-long context, multimodal understanding, and strong coding capabilities, making it suitable for complex reasoning, software development, knowledge analysis, and agent automation tasks.