ChatBAZ Coder
AI API Gateway · Flagship coding models

Four flagship coding models.
One key.

ChatBAZ Coder is a drop-in gateway. Point your OpenAI or Anthropic SDK at a single endpoint and get streaming, vision, tool use and reasoning across Qwen3.6-Plus, GLM-5.1, Kimi-K2.6 and MiniMax-M2.7 — with per-key quotas and daily resets you never have to think about again.

Quick start Admin sign in
The Lineup

Four minds, one endpoint.

Every model below is routed through the same OpenAI & Anthropic-compatible surface — streaming, vision, tool use and thinking, translated on the wire.

Model

Qwen3.6-Plus

Agentic coding

Qwen's newest release. Enhanced agentic reliability — designed to chain tool calls without drifting off task.

Call as qwen3.6-plus
Model

GLM-5.1

Deep reasoning

Zhipu's flagship. Long-horizon engineering tasks executed autonomously, with visible thinking blocks.

Call as glm-5.1
Model

Kimi-K2.6

SOTA coding · swarms

Kimi's latest delivers SOTA-level code generation and agent-swarm coordination out of the box.

Call as kimi-k2.6
Model

MiniMax-M2.7

Self-evolving

MiniMax kicks off model self-evolution — iteratively builds its own harness for sophisticated productivity tasks.

Call as minimax-m2.7
What you get

Everything your agent already speaks.

Streaming, vision, tool-calls and thinking deltas — all translated on the wire so your SDK doesn't have to change a single line.

⤼

Streaming

SSE chunks in the exact shape OpenAI & Anthropic SDKs expect.

◉

Vision

Pass images as image_url data URIs or image.source — routed natively.

◎

Tool use

Model emits tool calls; your agent runtime takes it from there.

✦

Thinking

Anthropic thinking blocks are forwarded when the model reasons.

⟢

OpenAI SDK

/v1/chat/completions and /v1/responses end-points.

⟣

Anthropic SDK

/v1/messages with system, tools, vision, thinking.

✶

Daily quotas

Per-key limits reset at 00:00 Europe/Istanbul — no silent surprises.

Quick start

Swap one line. Keep your SDK.

Point base_url at https://coder.chatbaz.app and your existing code Just Works. 

from openai import OpenAI

client = OpenAI(
    base_url="https://coder.chatbaz.app/v1",
    api_key="YOUR_KEY",
)

stream = client.chat.completions.create(
    model="qwen3.6-plus",
    messages=[{"role": "user", "content": "hello"}],
    stream=True,
)
for chunk in stream:
    print(chunk.choices[0].delta.content or "", end="")
from anthropic import Anthropic

client = Anthropic(
    base_url="https://coder.chatbaz.app",
    api_key="YOUR_KEY",
)

with client.messages.stream(
    model="glm-5.1",
    max_tokens=1024,
    messages=[{"role": "user", "content": "Explain quicksort."}],
) as stream:
    for text in stream.text_stream:
        print(text, end="")
import httpx

r = httpx.post(
    "https://coder.chatbaz.app/v1/responses",
    headers={"Authorization": "Bearer YOUR_KEY"},
    json={
        "model": "minimax-m2.7",
        "input": "Write a haiku about compilers.",
    },
)
print(r.json()["output"][0]["content"][0]["text"])
curl https://coder.chatbaz.app/v1/chat/completions \
  -H "Authorization: Bearer $KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "kimi-k2.6",
    "messages": [{"role":"user","content":"Say hello"}]
  }'

Have a key already? Check its remaining days and quota →