Integrations
19 tools, IDEs, agent runners, SDKs, and frameworks confirmed compatible. Every entry has a copy-paste setup snippet — no pseudo-code, no "should work" hedging.
Anthropic's official terminal coding agent. Drop-in: one env var change keeps Opus / Sonnet / Haiku routing.
export ANTHROPIC_BASE_URL="https://api.anvat.app"
export ANTHROPIC_API_KEY="sk-anvat-..."
# Use claude-code as normal
claude-code "refactor my login flow"AI-native IDE. Configure Anvat as an OpenAI-compatible custom endpoint to route Cursor's calls.
# Cursor → Settings → Models → OpenAI API Key
API Key: sk-anvat-...
Override OpenAI Base URL: https://api.anvat.app/v1
# Then add any custom model id (e.g. claude-sonnet-4-6)High-performance code editor with an Anthropic + OpenAI provider model. Anvat slots in as either.
// settings.json
{
"assistant": {
"default_model": { "provider": "anthropic", "model": "claude-sonnet-4-6" },
"providers": {
"anthropic": {
"api_url": "https://api.anvat.app",
"api_key": "sk-anvat-..."
}
}
}
}Open-source VS Code / JetBrains AI assistant. Configure Anvat as a custom Anthropic or OpenAI provider.
// ~/.continue/config.json
{
"models": [
{
"title": "Anvat — Sonnet",
"provider": "anthropic",
"model": "claude-sonnet-4-6",
"apiBase": "https://api.anvat.app",
"apiKey": "sk-anvat-..."
}
]
}Autonomous coding agent inside VS Code. Routes Anthropic calls; Anvat slots in transparently.
# VS Code → Cline → Settings
Anthropic API Key: sk-anvat-...
Anthropic Base URL: https://api.anvat.appCline fork with extended agentic features. Same Anthropic provider config.
# Roo settings → Anthropic
Base URL: https://api.anvat.app
API Key: sk-anvat-...Terminal pair-programmer. Use as Anthropic or OpenAI-compatible via env vars.
export ANTHROPIC_API_BASE="https://api.anvat.app"
export ANTHROPIC_API_KEY="sk-anvat-..."
aider --model anthropic/claude-sonnet-4-6 ./srcThe official Python OpenAI SDK works against Anvat with a base_url override.
from openai import OpenAI
client = OpenAI(
api_key="sk-anvat-...",
base_url="https://api.anvat.app/v1",
)
res = client.chat.completions.create(
model="claude-sonnet-4-6",
messages=[{"role": "user", "content": "hi"}],
)Official Node / browser OpenAI SDK with the same base_url override.
import OpenAI from "openai";
const client = new OpenAI({
apiKey: "sk-anvat-...",
baseURL: "https://api.anvat.app/v1",
});
const res = await client.chat.completions.create({
model: "gpt-5",
messages: [{ role: "user", content: "hi" }],
});Official Anthropic SDK with a base_url override. Streaming, tool use, caching all work.
import anthropic
client = anthropic.Anthropic(
api_key="sk-anvat-...",
base_url="https://api.anvat.app",
)
res = client.messages.create(
model="claude-opus-4-8",
max_tokens=1024,
messages=[{"role": "user", "content": "hi"}],
)Official @anthropic-ai/sdk with base URL override. Identical streaming + caching shape.
import Anthropic from "@anthropic-ai/sdk";
const client = new Anthropic({
apiKey: "sk-anvat-...",
baseURL: "https://api.anvat.app",
});
const res = await client.messages.create({
model: "claude-sonnet-4-6",
max_tokens: 1024,
messages: [{ role: "user", content: "hi" }],
});ChatOpenAI and ChatAnthropic both honor a base_url override — point them at Anvat.
from langchain_openai import ChatOpenAI
llm = ChatOpenAI(
model="claude-sonnet-4-6",
openai_api_base="https://api.anvat.app/v1",
openai_api_key="sk-anvat-...",
)
llm.invoke("hello")OpenAI / Anthropic LLM modules accept api_base — Anvat plugs in as a wholesale LLM provider for RAG and agents.
from llama_index.llms.openai_like import OpenAILike
llm = OpenAILike(
api_base="https://api.anvat.app/v1",
api_key="sk-anvat-...",
model="claude-sonnet-4-6",
is_chat_model=True,
)Self-hosted LiteLLM gateway can use Anvat as an upstream provider — best of both worlds for teams already on LiteLLM.
# litellm config.yaml
model_list:
- model_name: anvat-sonnet
litellm_params:
model: openai/claude-sonnet-4-6
api_base: https://api.anvat.app/v1
api_key: sk-anvat-...Compile prompts into reliable pipelines. DSPy's LM class accepts custom api_base.
import dspy
lm = dspy.LM(
model="openai/claude-sonnet-4-6",
api_base="https://api.anvat.app/v1",
api_key="sk-anvat-...",
)
dspy.configure(lm=lm)Use @ai-sdk/openai or @ai-sdk/anthropic with a baseURL pointing at Anvat.
import { createOpenAI } from "@ai-sdk/openai";
const anvat = createOpenAI({
baseURL: "https://api.anvat.app/v1",
apiKey: "sk-anvat-...",
});
await generateText({ model: anvat("claude-sonnet-4-6"), prompt: "hi" });Anything that thinks it's calling OpenRouter via its OpenAI-compatible endpoint will work against Anvat by swapping the base URL.
# Replace https://openrouter.ai/api/v1 with:
https://api.anvat.app/v1Open-Webui, AnythingLLM, Khoj, BigAGI, LibreChat, Lobe Chat — every one of these has an OpenAI-compatible endpoint field. Point at Anvat.
OpenAI base URL: https://api.anvat.app/v1
OpenAI API key: sk-anvat-...Open-source LLM eval framework. Configure Anvat as an OpenAI-compatible provider for cost-effective eval runs.
# promptfooconfig.yaml
providers:
- id: openai:chat:claude-sonnet-4-6
config:
apiBaseUrl: https://api.anvat.app/v1
apiKey: sk-anvat-...If a tool sets OPENAI_BASE_URL or ANTHROPIC_BASE_URL (or accepts a custom OpenAI / Anthropic provider URL), it works on Anvat — even if we haven't listed it. Email hello@anvat.app and we'll add it (and test it on our side) within a day or two.