Generative AI
Use models through a familiar API, create keys in the console, set a budget, and follow Brazilian-real consumption alongside your applications and databases.
Compare pricing, context, and capabilities for every available model.
Intelligence Index by Artificial Analysis · version v4.3 · data retrieved on 9/15/2026
The AI experience follows the same platform logic: create a key, integrate with a base URL, measure consumption, and evolve usage predictably.
Review available models, context window, concurrency, and consumption limits before taking a workload to production. The technical documentation keeps parameters and errors current for more sensitive integrations.
Cases where keeping the AI API, budget, and consumption in the same console as applications and databases makes sense.
Use it when you want to integrate AI into the product without separating infrastructure, keys, budgets, and consumption into another console.
Connect internal systems, support, and automations through a familiar API.
Extract, classify, and transform text while following Brazilian-real consumption.
A short sequence to create a key, change the base URL, and observe consumption before scaling usage.
Create an API key per project, with a budget to limit use and reduce surprise consumption.
Use OpenAI compatibility with Zenifra’s base_url and the model available in the console.
Follow calls, costs, and limits before expanding to critical flows.
Decisions that help separate environments, limit budgets, and understand current API limits.
Choose the model available for the use case and confirm current limitations.
Consumption control per key helps separate environments and integrations.
Check the context window and limits before more sensitive workloads.
Use separate keys for development, staging, and production.
Test with the same interface you already use in OpenAI-compatible projects: change the base URL, use your Zenifra key, and send the first request.
curl https://ai.zenifra.com/v1/chat/completions \
-H "Authorization: Bearer $ZENIFRA_AI_API_KEY" \
-H "Content-Type: application/json" \
-d '{"model":"zenifra/deepseek-v4-flash-0731","messages":[{"role":"user","content":"Hello, who are you?"}]}' \
| jq -r '.choices[0].message.content'import OpenAI from 'openai'
const client = new OpenAI({
apiKey: process.env.ZENIFRA_AI_API_KEY,
baseURL: 'https://ai.zenifra.com/v1',
})
const completion = await client.chat.completions.create({
model: 'zenifra/deepseek-v4-flash-0731',
messages: [
{ role: 'user', content: 'Hello, who are you?' },
],
})
console.log(completion.choices[0].message.content)import os
from openai import OpenAI
client = OpenAI(
api_key=os.environ["ZENIFRA_AI_API_KEY"],
base_url="https://ai.zenifra.com/v1",
)
completion = client.chat.completions.create(
model="zenifra/deepseek-v4-flash-0731",
messages=[
{"role": "user", "content": "Hello, who are you?"},
],
)
print(completion.choices[0].message.content)Create per-project keys, define a budget per key, and follow consumption to keep predictability before scaling usage.
Short answers to decide whether the AI API already fits your workflow.
Create keys per project, define a budget per key, and follow use in the console before increasing volume.
In most cases, use an OpenAI-compatible client and change the base URL, model, and API key.