Documentation

Quickstart

Create an API key in your dashboard, then make a request. Every new account starts with $5 in free credit, so you can run this immediately.

curl https://api.textclf.com/v1/chat/completions \
  -H "Authorization: Bearer $TEXTCLF_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "meta-llama/Llama-3.1-8B-Instruct",
    "messages": [{ "role": "user", "content": "Hello" }]
  }'

from openai import OpenAI

client = OpenAI(
    base_url="https://api.textclf.com/v1",
    api_key="YOUR_TEXTCLF_API_KEY",
)

resp = client.chat.completions.create(
    model="meta-llama/Llama-3.1-8B-Instruct",
    messages=[{"role": "user", "content": "Hello"}],
)
print(resp.choices[0].message.content)

import OpenAI from "openai"

const client = new OpenAI({
  baseURL: "https://api.textclf.com/v1",
  apiKey: process.env.TEXTCLF_API_KEY,
})

const resp = await client.chat.completions.create({
  model: "meta-llama/Llama-3.1-8B-Instruct",
  messages: [{ role: "user", content: "Hello" }],
})
console.log(resp.choices[0].message.content)

OpenAI compatibility

Every model is served behind an OpenAI-compatible API (the same interface exposed by vllm serve), so anything that speaks the OpenAI protocol works by changing two settings: the base URL and the API key. No SDK forks, no request rewriting.

• OpenAI Python SDKbase_url=
• OpenAI Node / TypeScript SDKbaseURL:
• Vercel AI SDK (@ai-sdk/openai)createOpenAI({ baseURL })
• LangChain (ChatOpenAI)openai_api_base=
• LlamaIndex (OpenAI)api_base=

Standard sampling parameters pass straight through to vLLM — temperature, top_p, max_tokens, stop, stream, and tool/function calling on models that support it.

# TextCLF speaks the OpenAI protocol, so point any
# OpenAI-compatible client at the base URL and key.

# OpenAI Python / TypeScript SDK  -> set base_url / baseURL
# LangChain (ChatOpenAI)          -> set openai_api_base
# LlamaIndex (OpenAI)             -> set api_base
# Vercel AI SDK (@ai-sdk/openai)  -> createOpenAI({ baseURL })

from openai import OpenAI
client = OpenAI(base_url="https://api.textclf.com/v1", api_key="YOUR_TEXTCLF_API_KEY")

Authentication

TextCLF keys begin with sk-tclf- and are passed as a bearer token on every request. Keys are shown once at creation — store them somewhere safe. You can create, name, and revoke keys at any time from the dashboard.

Authorization: Bearer sk-tclf-...

Never expose a key in client-side code or commit it to version control. If a key leaks, revoke it — usage stops immediately.

Endpoints

The core endpoint is chat completions. It accepts the same parameters as OpenAI, so the official openai SDKs work without modification.

• POST/v1/chat/completionsStreaming and non-streaming
• GET/v1/modelsList available models

Streaming

Set stream: true to receive server-sent events as tokens are generated. The final chunk carries the usage object, which is what we meter your credit against.

stream = client.chat.completions.create(
    model="meta-llama/Llama-3.1-8B-Instruct",
    messages=[{"role": "user", "content": "Hello"}],
    stream=True,
)
for chunk in stream:
    print(chunk.choices[0].delta.content or "", end="")

Errors

Errors use standard HTTP status codes with an OpenAI-shaped JSON body.

• 400Malformed request or unknown model ID.
• 401Missing or invalid API key.
• 402Insufficient credit — top up to continue.
• 429Rate limit exceeded — retry with backoff.
• 503Model temporarily unavailable — try another model.
• 504Upstream gateway timed out.

Models

Pass any of the model IDs below in the model field. See the full model catalog for pricing, context windows, and latency.

• meta-llama/Llama-3.3-70B-Instruct$0.08 / $0.30 per 1M
• meta-llama/Llama-3.1-8B-Instruct$0.018 / $0.038 per 1M
• Qwen/Qwen3.8-27B$0.38 / $2.98 per 1M

Billing & credit

TextCLF is prepaid. You hold a credit balance and each request deducts its exact cost — prompt tokens times the model's input rate, plus completion tokens times its output rate. There are no subscriptions and credit never expires.

• Free credit. Every new account gets $5, no card required.
• Top up any time. Add credit from the billing page in preset or custom amounts.
• Auto-refill. Set a threshold and amount, and we top up from your saved card before you run dry.