Tara in your editor
Use Tara in coding assistants and editors. Tara answers in the two formats most tools already speak: the OpenAI Chat Completions format and the Anthropic Messages format. If your tool lets you set a custom base URL and an API key, it can use Tara.
Get started
- Apply for a free Tara API key. A person reviews every application, usually within a couple of business days.
- Once approved, sign in and create a key on your account page. It starts with
tara_sk_and is shown once. - Point your tool at Tara with one of the setups below, and choose the model
tara.
Endpoints
| Format | Base URL | Endpoint | Key goes in |
|---|---|---|---|
| OpenAI compatible | https://tara.tratok.com/v1 | POST /v1/chat/completionsGET /v1/models | Authorization: Bearer tara_sk_... |
| Anthropic compatible | https://tara.tratok.com | POST /v1/messagesPOST /v1/messages/count_tokens | x-api-key: tara_sk_... |
The model name is tara. Other model names are accepted too, so a tool with a fixed list still works; the answer always comes from Tara.
Works: streaming, tool use (function calling), system prompts, stop sequences, JSON replies on request.
Not yet: images and files, audio, embeddings, more than one choice per request.
opencode
Add Tara as a provider in opencode.json (in your project, or ~/.config/opencode/opencode.json), set TARA_API_KEY in your environment, then pick Tara with /models.
{
"$schema": "https://opencode.ai/config.json",
"provider": {
"tara": {
"npm": "@ai-sdk/openai-compatible",
"name": "Tara by Tratok",
"options": {
"baseURL": "https://tara.tratok.com/v1",
"apiKey": "{env:TARA_API_KEY}"
},
"models": {
"tara": { "name": "Tara" }
}
}
}
}
Continue (VS Code and JetBrains)
Add a model to your Continue config.yaml:
models:
- name: Tara
provider: openai
model: tara
apiBase: https://tara.tratok.com/v1
apiKey: tara_sk_...
roles: [chat, edit, apply]
Cline
In Cline's settings choose the API provider OpenAI Compatible, then enter:
- Base URL:
https://tara.tratok.com/v1 - API key: your
tara_sk_...key - Model ID:
tara
Aider
export OPENAI_API_BASE=https://tara.tratok.com/v1
export OPENAI_API_KEY=tara_sk_...
aider --model openai/tara
Any tool with an Anthropic-compatible setting
Set the base URL to https://tara.tratok.com (without /v1: these tools add it), the API key to your Tara key, and the model to tara.
From your own code
OpenAI SDK (Python)
from openai import OpenAI
client = OpenAI(base_url="https://tara.tratok.com/v1", api_key="tara_sk_...")
reply = client.chat.completions.create(
model="tara",
messages=[{"role": "user", "content": "Write a haiku about the sea."}],
)
print(reply.choices[0].message.content) OpenAI SDK (JavaScript)
import OpenAI from "openai";
const client = new OpenAI({ baseURL: "https://tara.tratok.com/v1", apiKey: process.env.TARA_API_KEY });
const stream = await client.chat.completions.create({
model: "tara",
messages: [{ role: "user", content: "Explain recursion in two sentences." }],
stream: true,
});
for await (const chunk of stream) process.stdout.write(chunk.choices[0]?.delta?.content ?? ""); Anthropic SDK (Python)
import anthropic
client = anthropic.Anthropic(base_url="https://tara.tratok.com", api_key="tara_sk_...")
message = client.messages.create(
model="tara",
max_tokens=1024,
messages=[{"role": "user", "content": "Suggest three names for a coffee shop."}],
)
print(message.content[0].text) curl
curl https://tara.tratok.com/v1/chat/completions \
-H "Authorization: Bearer $TARA_API_KEY" \
-H "Content-Type: application/json" \
-d '{"model": "tara", "stream": true, "messages": [{"role": "user", "content": "Hello, Tara"}]}'
Limits
- Each account starts with 60 requests a minute, 1,000 requests a day and 100,000 tokens a day. The daily limits reset at 00:00 UTC.
- Coding assistants send the whole conversation, often with files, at every step, so they use far more tokens than a chat app. If you hit the daily limit, ask for more by replying to your approval email.
- A reply can be up to 8,192 tokens, and a request up to 4 MB.
- When a limit is reached the API answers HTTP 429 with a
Retry-Afterheader.
Privacy
Tara does not store what you send through the API. We keep only usage counts (requests, tokens, timing) for your account and its limits. See the privacy page.