Enverth Conduit · Live

Every open model. One endpoint.

Fast, OpenAI-compatible access to the best open-weight models with zero data retention, for up to 75% less than official API prices.

Get your API keyBrowse models

From $0.09 / 1M tokens · ZDR · No minimums

PromptTokens →

■ Quickstart

One base URL.

Conduit speaks the OpenAI API. Point any OpenAI SDK at Conduit, pick a model and you're live. No new client library, no migration.

Serving DeepSeek, Z.ai, Xiaomi models, from $0.09 per million input tokens.

Read the quickstart
from openai import OpenAI

client = OpenAI(
    base_url="https://api.conduit.enverth.ai/v1",
    api_key="CONDUIT_API_KEY",
)

response = client.chat.completions.create(
    model="deepseek/deepseek-v4.1-flash",
    messages=[{"role": "user", "content": "Explain MoE in one sentence."}],
)

print(response.choices[0].message.content)

■ Why Conduit

Inference without the infrastructure.

01

OpenAI-compatible

Swap one base URL and keep your existing SDKs, agents and tooling. No rewrites.

02

Pay per token

No subscriptions, no minimums, no idle GPUs. You pay only for tokens in and out.

03

Zero data retention

Every model runs with ZDR. Prompts and completions are never stored or used for training.

04

Low latency

Optimized serving stacks with speculative decoding and fast time-to-first-token.

05

Scales with you

From your first prototype to billions of tokens a day, with no capacity planning.

06

Streaming & tool calls

Server-sent events, function calling and JSON mode on every model that supports them.

■ Models

The frontier of open weights.

001 / DeepSeek1M context

DeepSeek V4.1 Flash

Fast, low-cost MoE with native image input and a 1M-token context.

ZDRSave up to 75%
Input / 1M
$0.09$0.30 −70%
Output / 1M
$0.30$1.20 −75%
Cache read
$0.006 

deepseek/deepseek-v4.1-flash

002 / Z.ai1M context

GLM 5.3

Z.ai's flagship for agentic coding and long-horizon reasoning.

ZDRSave up to 75%
Input / 1M
$0.60$1.40 −57%
Output / 1M
$1.10$4.40 −75%
Cache read
$0.20$0.26 −23%

zai/glm-5.3

003 / Xiaomi1M context

MiMo V2.6 Pro

Trillion-parameter MoE built for agents, tool use and coding.

ZDR
Input / 1M
$0.43 
Output / 1M
$0.87 
Cache read
$0.0036 

xiaomi/mimo-v2.6-pro

004 / Z.ai1M context

GLM 5.3 Flash

Lightweight GLM for high-volume, low-latency workloads.

ZDRSave up to 40%
Input / 1M
$0.10$0.15 −33%
Output / 1M
$0.30$0.50 −40%
Cache read
$0.025$0.03 −17%

zai/glm-5.3-flash

005 / Z.ai1M context

GLM 5.3 Uncensored

GLM 5.3 without content refusals, for research and unrestricted creative work.

ZDRUncensored
Input / 1M
$1.125 
Output / 1M
$2.025 
Cache read
$0.27 

zai/glm-5.3-uncensored

006 / Z.ai1M context

GLM 5.3 Flash Uncensored

GLM 5.3 Flash without content refusals, at Flash speed and cost.

ZDRUncensored
Input / 1M
$0.18 
Output / 1M
$0.72 
Cache read
$0.063 

zai/glm-5.3-flash-uncensored

All 6 models and pricing →

■ How it works

Live in a minute.

  1. Step 01

    Create a key

    Sign up and generate an API key in seconds.

  2. Step 02

    Point your SDK

    Set the base URL to Conduit. Your OpenAI client works as-is.

  3. Step 03

    Pick any model

    Switch between DeepSeek, GLM and MiMo by changing one string.

■ Start

Start building.

Pay-as-you-go from $0.09 per million tokens, with zero data retention on every model.

Get your API keyRead the docs

Volume pricing or questions: [email protected]