api.jatevo.ai ยท OpenAI-compatible

One endpoint.
Many upstreams.

Use one Jatevo API key and route requests to Kimi K2.7 Code, Cerebras Gemma, LB/Codex, GLM, Qwen, NVIDIA Nemotron, or any configured model provider. The upstream vendor keys stay inside Jatevo's server-side pool.

from openai import OpenAI

client = OpenAI(
  base_url="https://api.jatevo.ai/v1",
  api_key="sk-jvo-..."
)

resp = client.chat.completions.create(
  model="kimi-k2.7-code",
  messages=[{"role":"user","content":"Write one sentence about Jatevo."}],
  stream=True
)

for chunk in resp:
  if chunk.choices and chunk.choices[0].delta.content:
    print(chunk.choices[0].delta.content, end="")
/v1
standard base path
OpenAI-compatible clients
GPT
3.jatevo upstream
gpt-5.6-sol, terra, 5.5
glm/
GLM route
glm/GLM-5.1 etc
cerebras/
Gemma route
cerebras/gemma-4-31b
kimi
Kimi K2.7 Code
kimi-k2.7-code
nvidia/
Nemotron route
nvidia/Nemotron-3-Ultra...
1 key
user auth layer
vendor keys hidden
Routing

How model routing works

Preferred routes: send requests to https://api.jatevo.ai/v1 and set model="kimi-k2.7-code" for Kimi or model="cerebras/gemma-4-31b" for Cerebras Gemma. Chat Completions, /v1/responses where supported, function tools where supported, and OpenAI-compatible stream=True SSE are supported.