x0root AI API - Monochrome Playground

Model:

x0root AI API Playground

Monochrome OpenAI-compatible client. Features intelligent model routing, NVIDIA NIM waterfalls, Puter fallback, and Durable Object memory.

Chat: default Shift + Enter for new line

Supported Models & Aliases

List of available models, routing aliases, and Zen provider variants.

router / auto

Full auto-routing: classifies prompt into execution tiers (LITE to HEAVY) and tries NVIDIA NIM then Puter waterfall.

zen / zen-auto

Forces Zen provider by selecting a random Zen model from the classified tier.

puter / puter-auto

Forces Puter provider directly, skipping NVIDIA NIM.

nvidia / nvidia-auto

Forces NVIDIA NIM provider only, with Puter fallback.

deepseek-v4-flash-free

Available Zen model (supports variants: low, medium, high, max).

mimo-v2.5-free

Available Zen model (supports variants: low, medium, high).

big-pickle

Available Zen model (no variants).

nemotron-3-ultra-free

Available Zen model (supports variants: low, medium, high).

API Explorer

Test endpoints against `puter-api.x0root.workers.dev`.

GET /

GET /v1/models

POST /v1/chat/completions

Router Architecture

Overview of classification tiers and NVIDIA/Puter fallback.

LITE

Casual chat, short answers, extraction.

NORMAL

Daily assistant tasks, coding help.

NORMAL_CALCULATION

Math, physics, algorithms.

MEDIUM

Nuanced writing, heavy coding.

HEAVY_CALCULATION

Advanced proofs & derivations.

HEAVY

Complex reasoning & debugging.

Settings & Auth

Configure your API key and Worker URL.

Worker Base URL

API Key (`qp-...`)

x0root AI API Playground

Supported Models & Aliases

API Explorer

Router Architecture

Settings & Auth

API Key Required