J JJAPI
δΈ­ζ–‡

xAI

Call xAI's Grok API without an X premium account.

grok-2, grok-2-mini, grok-2-vision β€” xAI's models with realtime knowledge and unfiltered output, through one OpenAI-compatible endpoint.

Get my API key β†’ Quickstart in 30 seconds β†’

Latest models as of 2026-05-23

See all models β†’
grok-4 flagship

xAI's current flagship with realtime knowledge access via X integration.
grok-4-fast fast

Cheaper, faster Grok 4 variant. For chatbot use cases.
grok-3 balanced

Prior-gen Grok. Available for legacy compatibility.

Why use JJAPI for this API

No X Premium needed

Direct Grok API access without paying for X Premium+ or applying for the xAI waitlist.

Realtime knowledge

Grok has tighter integration with realtime sources than most LLMs β€” useful for news, sports, and current events.

OpenAI-compatible

Same SDK signature β€” switch between Grok and GPT-4o by changing one string.

Higher tolerance

Grok has a more permissive content policy than GPT-4 / Claude for legitimate creative work.

example.py 
from openai import OpenAI

client = OpenAI(
    base_url="https://api.jjapi.net/v1",
    api_key="sk-jjapi-...",
)

r = client.chat.completions.create(
    model="grok-4",  # current as of 2026-05-23
    messages=[{"role": "user", "content": "Hi"}],
)
print(r.choices[0].message.content)

Frequently asked

Does Grok have realtime web access?

Yes β€” Grok can access realtime information from X and the broader web during inference. Latency is slightly higher than non-grounded models.

Is Grok cheaper than GPT-4o?

Grok-2 is similar to GPT-4o pricing. Grok-2-mini is ~50% cheaper than GPT-4o-mini for most tasks.

Does function calling work?

Yes β€” Grok supports OpenAI-style 'tools' parameter with parallel tool calls.

Get my API key β†’