Pull ChatGPT answers as structured data.
Submit a prompt, get a parsed response: answer text, citations, code blocks, and tables, all returned as typed JSON. We handle the session, the throttling, and the model selection.
GET /chatgpt
?prompt=compare+GPT-5+and+Claude+4.5
→ 200 OK · 15 credits · 3.1s
{
"recipe": "ChatGPT",
"conversation": {
"messages": [
{ "role": "user", "content": "compare GPT-5 and Claude 4.5" },
{ "role": "assistant", "content": "GPT-5 leads on SWE-bench (74%)..." }
],
"total_messages": 2,
"user_messages": 1,
"assistant_messages": 1
}
}See it work, before you sign up.
Drop in a URL, run a real call against the live API, and watch the JSON come back in about a second. No API key required.
gpt-5, gpt-4o, o3, o3-mini, o1, gpt-4-turbo. We track new releases and route requests; you just pick the family.
Set format:"structured" and we split the response into answer text, citations, code blocks, and tables. No regex. No HTML tag-soup.
Web-search-enabled responses come back with the source URLs ChatGPT actually consulted, not just the bracketed [1][2] markers in the prose.
Set stream:true and tokens flow back as they arrive. Identical interface to OpenAI's native stream, drop-in for existing UIs.
Pass thread_id to continue a session. We persist context server-side; no need to ship the full history with every call.
You don't need an OpenAI account, login, or your own quota. We pool capacity across our infrastructure and bill per-success only.
Copy. Paste. Ship.
import { Ujeebu } from "ujeebu";
const uj = new Ujeebu(process.env.UJEEBU_KEY);
const { data } = await uj.chatgpt({
prompt: "compare GPT-5 and Claude 4.5 on coding",
model: "gpt-5",
format: "structured",
});
console.log(data.answer);
data.citations.forEach(c => console.log(c.url));from ujeebu import Ujeebu
uj = Ujeebu(api_key=os.environ["UJEEBU_KEY"])
result = uj.chatgpt(
prompt="compare GPT-5 and Claude 4.5 on coding",
model="gpt-5",
format="structured",
)
print(result["answer"])
for c in result["citations"]: print(c["url"])curl -G https://api.ujeebu.com/chatgpt \
-H "ApiKey: $UJEEBU_KEY" \
--data-urlencode "prompt=compare GPT-5 and Claude 4.5"const stream = await uj.chatgpt({
prompt: "write a 500-word essay on...",
model: "gpt-5",
stream: true,
});
for await (const chunk of stream) {
process.stdout.write(chunk.delta);
}Real things real teams shipped this quarter.
Track how ChatGPT answers your brand and competitor queries. Daily cron, diff against yesterday, alert on shifts in cited sources.
Generative Engine Optimization. Run thousands of buyer-intent prompts daily and measure which sources ChatGPT pulls into its answers.
Run benchmark prompts at scale across gpt-5/4o/o3 in parallel. Concurrency and retry are ours; you get clean comparison rows.
Power chat features inside your product without managing OpenAI keys, quotas, or model deprecations. Per-success billing matches your usage.
15 credits per ChatGPT Scraper call. Here's what that buys.
One credit pool covers every endpoint. Failed calls cost 0. No per-feature upcharges, no premium-proxy tax. See full pricing →
Ship ChatGPT Scraper tonight.
5,000 credits free. No card. Real residential proxies on the free tier.