AICodeDog

One-click integration

Change one Base URL line —
14+ mainstream coding tools work instantly.

We implement both OpenAI-compatible (/v1/chat/completions) and Anthropic-native (/v1/messages) protocols, so whichever camp your tool belongs to, it just works.

OpenAI-compatible

https://aicodedog.com/v1

Works with: Cursor · Aider · OpenAI SDK · OpenHands · Codex · Continue (OpenAI mode)

Anthropic-native

https://aicodedog.com

Works with: Claude Code · Cline · Roo · OpenCode · Anthropic SDK

Get my API KeyFull docsAvailable models

CLI tools6 tools

Command-line first — best for agent automation and server deployments.

Claude Code

Anthropic

Anthropic's official coding CLI — AICodeDog's #1 compatibility target

Website
Paste your key to make every snippet ready to run:

Add these two env vars to your shell rc (~/.zshrc / ~/.bashrc); restart your terminal and every `claude` command will route through AICodeDog.

~/.zshrc · bash
export ANTHROPIC_BASE_URL="https://aicodedog.com"
export ANTHROPIC_AUTH_TOKEN="<YOUR_API_KEY>"
Verify · bash
claude code "write a python fibonacci function"

OpenAI Codex CLI

OpenAI compatible

OpenAI's official coding CLI — let GPT rewrite your repo

Website
Paste your key to make every snippet ready to run:
~/.zshrc · bash
export OPENAI_BASE_URL="https://aicodedog.com/v1"
export OPENAI_API_KEY="<YOUR_API_KEY>"
Verify · bash
codex "explain the logic on line 100 of src/main.ts"

Gemini CLI

Gemini · Beta

Google's official Gemini CLI (experimental — requires Google Vertex upstream)

Website
Paste your key to make every snippet ready to run:

AICodeDog's Gemini compat layer is currently in Beta — recommended for testing only. For production prefer Anthropic or OpenAI protocols.

bash
export GEMINI_BASE_URL="https://aicodedog.com/v1"
export GEMINI_API_KEY="<YOUR_API_KEY>"
gemini "hello"

Aider

Both protocols

Lightweight CLI — pair-programming with git, battle-tested

Website
Paste your key to make every snippet ready to run:
~/.zshrc · bash
export OPENAI_API_BASE="https://aicodedog.com/v1"
export OPENAI_API_KEY="<YOUR_API_KEY>"
Run · bash
aider --model openai/gpt-5.4 src/main.ts
# or use sonnet:
aider --model openai/claude-sonnet-4-6

OpenCode

Anthropic

One of cc-switch's 5 supported tools — fully open source

Website
Paste your key to make every snippet ready to run:
bash
export ANTHROPIC_BASE_URL="https://aicodedog.com"
export ANTHROPIC_API_KEY="<YOUR_API_KEY>"
opencode

OpenHands (formerly OpenDevin)

OpenAI compatible

Fully autonomous agent — opens its own browser to run code

Website
Paste your key to make every snippet ready to run:

After launching the OpenHands UI, go to Settings → choose Custom LLM →

bash
Provider:    OpenAI
Base URL:    https://aicodedog.com/v1
API Key:     <YOUR_API_KEY>
Model:       openai/gpt-5.4

Editors / IDEs2 tools

Embed an LLM into your editor — closest to the "Cursor-like" experience.

Cursor

OpenAI compatible

The most popular AI-first editor — supports custom OpenAI-compatible endpoints

Website
Paste your key to make every snippet ready to run:

Settings → Models → scroll to the OpenAI API Key section. Toggle on "Override OpenAI Base URL".

The two fields to fill · json
Base URL:  https://aicodedog.com/v1
API Key:   <YOUR_API_KEY>

Click "Verify" — you should see a green checkmark. Then under "Add model", add the slugs you want to use, e.g. claude-sonnet-4-6, gpt-5.4.

Windsurf

OpenAI compatible

Codeium's Cascade IDE — native OpenAI protocol support

Website
Paste your key to make every snippet ready to run:

Settings → Cascade → Custom Model Provider → add OpenAI-Compatible.

json
{
  "provider": "openai",
  "baseUrl":  "https://aicodedog.com/v1",
  "apiKey":   "<YOUR_API_KEY>",
  "modelName":"gpt-5.4"
}

VS Code extensions3 tools

Plug into your existing VS Code — zero migration cost.

Cline

Both protocols

Auto-agent inside VS Code — edits files and runs commands on its own

Website
Paste your key to make every snippet ready to run:

Cline settings → set API Provider to "Anthropic" → enable "Use custom base URL".

bash
Base URL:  https://aicodedog.com
API Key:   <YOUR_API_KEY>
Model:     claude-sonnet-4-6

Or set API Provider to "OpenAI Compatible" → Base URL = ${BASE}/v1, Model = gpt-5.4. Both modes verified.

Continue.dev

Both protocols

Works in VS Code & JetBrains — config-file driven, can mount multiple providers at once

Website
Paste your key to make every snippet ready to run:

Edit ~/.continue/config.yaml (or .json), add to the `models` section:

~/.continue/config.yaml · yaml
models:
  - name: AICodeDog · Sonnet
    provider: anthropic
    apiBase: https://aicodedog.com
    apiKey: <YOUR_API_KEY>
    model: claude-sonnet-4-6
  - name: AICodeDog · GPT-5.4
    provider: openai
    apiBase: https://aicodedog.com/v1
    apiKey: <YOUR_API_KEY>
    model: gpt-5.4

Roo Code

Both protocols

Cline fork — more model sources + subtask scheduling

Website
Paste your key to make every snippet ready to run:

API Configuration → Provider "Anthropic" → enable Custom Base URL → fill in:

bash
Base URL:    https://aicodedog.com
API Key:     <YOUR_API_KEY>
Model:       claude-sonnet-4-6

SDKs2 tools

Call directly from your code — first choice for product integrations.

OpenAI SDK (Node / Python)

OpenAI compatible

Official SDK with the broadest coverage. Just change baseURL.

Website
Paste your key to make every snippet ready to run:
Node / TypeScript · ts
import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://aicodedog.com/v1",
  apiKey:  "<YOUR_API_KEY>",
});

const r = await client.chat.completions.create({
  model: "gpt-5.4",
  messages: [{ role: "user", content: "Hi" }],
});
Python · py
from openai import OpenAI

client = OpenAI(base_url="https://aicodedog.com/v1", api_key="<YOUR_API_KEY>")
r = client.chat.completions.create(
    model="gpt-5.4",
    messages=[{"role": "user", "content": "Hi"}],
)

Anthropic SDK

Anthropic

Official SDK — full Anthropic protocol (messages / tools / vision)

Website
Paste your key to make every snippet ready to run:
Node / TypeScript · ts
import Anthropic from "@anthropic-ai/sdk";

const client = new Anthropic({
  baseURL: "https://aicodedog.com",
  apiKey:  "<YOUR_API_KEY>",
});

const r = await client.messages.create({
  model: "claude-sonnet-4-6",
  max_tokens: 1024,
  messages: [{ role: "user", content: "Hi" }],
});

Generic / curl1 tools

No tool dependency — pure HTTP, fastest way to verify.

curl / HTTP

Both protocols

Zero dependencies — perfect for scripting & debugging

Website
Paste your key to make every snippet ready to run:
OpenAI protocol · bash
curl https://aicodedog.com/v1/chat/completions \
  -H "Authorization: Bearer <YOUR_API_KEY>" \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-5.4","messages":[{"role":"user","content":"hi"}]}'
Anthropic protocol · bash
curl https://aicodedog.com/v1/messages \
  -H "x-api-key: <YOUR_API_KEY>" \
  -H "anthropic-version: 2023-06-01" \
  -H "Content-Type: application/json" \
  -d '{"model":"claude-sonnet-4-6","max_tokens":256,"messages":[{"role":"user","content":"hi"}]}'

Don't see your tool?

If your tool supports a custom OpenAI or Anthropic Base URL, AICodeDog will work. Email us what you use — it'll ship in the next release.

Contact us