OpenAI-compatible gateway

Ship a cleaner AI product stack without the usual surface-area mess.

BimiCode runs a tight black-theme developer surface for auth, keys, routing, usage, ads-based access, and model delivery. Your app keeps the OpenAI request shape while the platform handles the rest.

2Public model tiers
48hKey lifetime
1OpenAI request shape
Supabase-backed auth cookie sessions ads-based key renewal usage analytics fast model routing
bash
$ 
Platform

Built for an actual API business

The product flow covers signup, cookies, dashboard visibility, sponsor inventory, forced renewal, billing, and docs. It is not just a proxy in a box.

Access

Persistent sessions

Users stay signed in through cookies instead of getting bounced back to login constantly.

Monetization

Ads-based renewal

API keys rotate on a 2-day lifecycle and renew only after the sponsor completion flow.

Storage

Supabase-backed data

Users, usage, abuse logs, plan state, and renewal status now sit in Supabase instead of local SQLite.

Routing

Stable model surface

The public `deepseek-v4-*` labels stay stable while the backend manages provider routing privately.

Features

Small surface, clear behavior

The product stays focused on what matters in practice: onboarding, routing, usage control, and monetization.

Drop-in Compatible

Works with any OpenAI SDK. Change one line of code.

Instant API Keys

Sign up and get your key in under 10 seconds.

Abuse Protected

Rate limiting, anomaly detection, prompt shields.

Usage Dashboard

Track tokens, requests, and RPM in real time.

Stable Public Labels

Keep `deepseek-v4-flash` and `deepseek-v4-pro` in your app while routing changes upstream.

Operator Controls

Ad placements, renewal gating, and plan upgrades live inside the same product surface.

Models

Choose your model

Same request contract, different latency and reasoning profiles.

DeepSeek V4 Flash

Fast
SpeedFastest
BehaviorLower latency
Best forQuick tasks, chat, code generation
RPM (Free)5
RPM (Pro)70

DeepSeek V4 Pro

Powerful
SpeedOptimized
BehaviorStronger reasoning
Best forComplex analysis and longer tasks
RPM (Free)5
RPM (Pro)70
Pricing

Simple pricing

A free entry point with ads-based renewal, plus a simple Pro tier for higher throughput.

Free

$0/month
  • 5 requests per minute
  • Both models
  • Usage dashboard
  • 2-day key access via sponsor flow
Get Started
Quick Start

Works with any OpenAI SDK

Keep the same SDK flow. Change the base URL and send requests.

python
import openai

client = openai.OpenAI(
    base_url="https://bimicode.xyz/v1",
    api_key="sk-bm-your-api-key"
)

response = client.chat.completions.create(
    model="deepseek-v4-flash",
    messages=[{"role": "user", "content": "Hello!"}]
)
print(response.choices[0].message.content)