Skip to content
Docs
Get started

Integrations

SDKs and frameworks

There's no FullsendAI SDK to install: the official OpenAI and Anthropic SDKs, and every framework built on them, work by changing the base URL and key.
SDKBase URL
OpenAI SDKs and OpenAI-compatible frameworkshttps://fullsendai.com/v1
Anthropic SDKshttps://fullsendai.com
Vercel AI SDK's Anthropic providerhttps://fullsendai.com/v1

OpenAI and Anthropic SDKs#

# pip install openaiimport osfrom openai import OpenAIclient = OpenAI(base_url="https://fullsendai.com/v1", api_key=os.environ["FULLSEND_API_KEY"])

Both SDKs also read their usual environment variables: OPENAI_BASE_URL and OPENAI_API_KEY, or ANTHROPIC_BASE_URL and ANTHROPIC_API_KEY.

Vercel AI SDK#

// npm install ai @ai-sdk/openai-compatibleimport { createOpenAICompatible } from "@ai-sdk/openai-compatible";import { streamText } from "ai";const fullsend = createOpenAICompatible({ name: "fullsend", baseURL: "https://fullsendai.com/v1", apiKey: process.env.FULLSEND_API_KEY });const result = streamText({ model: fullsend("deepseek-v4.1-flash"), prompt: "Write a SQL migration that adds a created_at column" });for await (const text of result.textStream) process.stdout.write(text);

LangChain#

# pip install langchain-openaiimport osfrom langchain_openai import ChatOpenAIllm = ChatOpenAI(model="deepseek-v4.1-flash", base_url="https://fullsendai.com/v1", api_key=os.environ["FULLSEND_API_KEY"])print(llm.invoke("Explain Python's GIL in two sentences").content)

LlamaIndex#

Python
# pip install llama-index-llms-openai-likeimport osfrom llama_index.llms.openai_like import OpenAILikellm = OpenAILike(model="deepseek-v4.1-flash", api_base="https://fullsendai.com/v1", api_key=os.environ["FULLSEND_API_KEY"], is_chat_model=True, is_function_calling_model=True)print(llm.complete("Summarise the CAP theorem"))

LiteLLM#

Python
# pip install litellmimport osfrom litellm import completionreply = completion(    model="openai/deepseek-v4.1-flash",    api_base="https://fullsendai.com/v1",    api_key=os.environ["FULLSEND_API_KEY"],    messages=[{"role": "user", "content": "Write a haiku about race conditions"}],)

Anything else#

Any library or tool with an “OpenAI-compatible” or “custom base URL” option works: give it https://fullsendai.com/v1, your key and a model ID. Coding agents and editors are covered on Connect your agent. Every call, including failed ones, shows in Logs with the reason, which is the quickest way to see what a tool sent.

Stuck? Email [email protected]System statusChangelog