🌐 pi-zyloo-provider

48 models across 9 families via Zyloo

Claude, GPT, Gemini, Grok, Kimi, GLM, DeepSeek, Qwen & MiniMax — one OpenAI-compatible endpoint for pi.

Features

• One endpoint, every family — Anthropic Claude, OpenAI GPT, Google Gemini, xAI Grok, Moonshot Kimi, Zhipu GLM, DeepSeek, Qwen & MiniMax behind a single OpenAI-compatible /v1/chat/completions API
• Reasoning models — thinking mode with reasoning_effort / reasoning_content on Claude, GPT, Gemini, GLM, DeepSeek, Qwen & MiniMax
• Vision models — image input on Claude, GPT, Gemini, Grok, Kimi & MiniMax M3
• Tool use — function calling support
• Streaming — real-time token streaming
• Stale-While-Revalidate — zero-latency startup from the embedded catalog, with a background refresh against https://api.zyloo.io/v1/models

Endpoint note: api.zyloo.ai has no DNS A record. The live API is served at api.zyloo.io; https://zyloo.ai/v1/* 301-redirects to https://api.zyloo.io/v1/*.

Available Models

| Model | Provider | Context | Vision | Reasoning | Input $/M | Output $/M | |-------|---------|---------|--------|-----------|-----------|------------| | Claude Fable 5 | Anthropic | 1.0M | ✅ | ✅ | $10.00 | $50.00 | | Claude Opus 4.6 | Anthropic | 1.0M | ✅ | ✅ | $5.00 | $25.00 | | Claude Opus 4.6 Thinking | Anthropic | 1.0M | ✅ | ✅ | $5.00 | $25.00 | | Claude Opus 4.7 | Anthropic | 1.0M | ✅ | ✅ | $5.00 | $25.00 | | Claude Opus 4.7 Thinking | Anthropic | 1.0M | ✅ | ✅ | $5.00 | $25.00 | | Claude Opus 4.8 | Anthropic | 1.0M | ✅ | ✅ | $5.00 | $25.00 | | Claude Opus 4.8 Thinking | Anthropic | 1.0M | ✅ | ✅ | $5.00 | $25.00 | | Claude Opus 5 | Anthropic | 1.0M | ✅ | ✅ | $5.00 | $25.00 | | Claude Opus 5 Thinking | Anthropic | 1.0M | ✅ | ✅ | $5.00 | $25.00 | | Claude Sonnet 4.6 | Anthropic | 1.0M | ✅ | ✅ | $3.00 | $15.00 | | Claude Sonnet 5 | Anthropic | 1.0M | ✅ | ✅ | $3.00 | $15.00 | | DeepSeek V3.1 | DeepSeek | 164K | ❌ | ✅ | $0.27 | $0.95 | | DeepSeek V3.2 | DeepSeek | 164K | ❌ | ✅ | $0.23 | $0.34 | | DeepSeek V4 Flash | DeepSeek | 1.0M | ❌ | ✅ | $0.14 | $0.28 | | DeepSeek V4 Pro | DeepSeek | 1.0M | ❌ | ✅ | $0.43 | $0.87 | | Gemini 2.5 Flash | Google | 1.0M | ✅ | ✅ | $0.30 | $2.50 | | Gemini 3 Flash Preview | Google | 1.0M | ✅ | ✅ | $0.50 | $3.00 | | Gemini 3 Pro | Google | 1.0M | ✅ | ✅ | $2.00 | $12.00 | | Gemini 3 Pro Preview | Google | 1.0M | ✅ | ✅ | $2.00 | $12.00 | | Gemini 3.1 Flash Lite Preview | Google | 1.0M | ✅ | ✅ | $0.25 | $1.50 | | Gemini 3.1 Pro Preview | Google | 1.0M | ✅ | ✅ | $2.00 | $12.00 | | Gemini 3.5 Flash | Google | 1.0M | ✅ | ✅ | $1.50 | $9.00 | | Gemini 3.6 Flash | Google | 1.0M | ✅ | ✅ | $1.50 | $9.00 | | GLM-4.7 | Zhipu | 205K | ❌ | ✅ | Free | Free | | GLM-5 | Zhipu | 200K | ❌ | ✅ | Free | Free | | GLM-5.1 | Zhipu | 200K | ❌ | ✅ | Free | Free | | GLM-5.2 | Zhipu | 1.0M | ❌ | ✅ | Free | Free | | GPT-4.1 | OpenAI | 1.0M | ✅ | ❌ | $2.00 | $8.00 | | GPT-4o | OpenAI | 128K | ✅ | ❌ | $2.50 | $10.00 | | GPT-5.1 | OpenAI | 400K | ✅ | ✅ | $1.25 | $10.00 | | GPT-5.2 | OpenAI | 400K | ✅ | ✅ | $1.75 | $14.00 | | GPT-5.4 | OpenAI | 272K | ✅ | ✅ | $2.50 | $15.00 | | GPT-5.5 | OpenAI | 272K | ✅ | ✅ | $5.00 | $30.00 | | GPT-5.6 Luna | OpenAI | 272K | ✅ | ✅ | $5.00 | $30.00 | | GPT-5.6 Sol | OpenAI | 272K | ✅ | ✅ | $5.00 | $30.00 | | GPT-5.6 Terra | OpenAI | 272K | ✅ | ✅ | $5.00 | $30.00 | | Grok 4.3 | xAI | 1.0M | ✅ | ✅ | $1.25 | $2.50 | | Grok 4.5 | xAI | 1.0M | ✅ | ✅ | $1.25 | $2.50 | | Grok Build 0.1 | xAI | 256K | ✅ | ✅ | $1.00 | $2.00 | | Kimi K2 Thinking | Moonshot | 262K | ❌ | ✅ | $0.60 | $2.50 | | Kimi K2.5 | Moonshot | 262K | ✅ | ✅ | $0.60 | $3.00 | | Kimi K2.6 | Moonshot | 262K | ✅ | ✅ | $0.95 | $4.00 | | Kimi K2.7 Code | Moonshot | 262K | ✅ | ✅ | $0.95 | $4.00 | | Kimi K3 | Moonshot | 262K | ✅ | ✅ | $0.95 | $4.00 | | MiniMax M2.5 | MiniMax | 205K | ❌ | ✅ | $0.15 | $0.90 | | MiniMax M3 | MiniMax | 512K | ✅ | ✅ | $0.60 | $2.40 | | Qwen3 Max | Qwen | 262K | ❌ | ❌ | $0.78 | $3.90 | | Qwen3.7 Max | Qwen | 1.0M | ❌ | ✅ | $1.25 | $3.75 |

Pricing reflects each model's native upstream provider (sourced from pi's built-in model catalog) where available. Zyloo's /v1/models endpoint exposes no pricing, and Zyloo may apply its own rates or markup — verify against your Zyloo dashboard. Models not yet present in any upstream catalog (e.g. GPT-5.6, Claude Opus 5, Gemini 3.6 Flash, Gemini 3 Pro, Grok 4.5, Kimi K3, GLM-5) use estimates from their closest known sibling.

Installation

Option 1: Using pi install (Recommended)

Install directly from GitHub:

pi install https://github.com/monotykamary/pi-zyloo-provider

Then set your API key and run pi:

# Recommended: add to auth.json
# See Authentication section below

# Or set as environment variable
export ZYLOO_API_KEY=your-api-key-here

pi

Get your API key from zyloo.io.

Option 2: Manual Clone

1. Clone this repository:

   git clone https://github.com/monotykamary/pi-zyloo-provider.git
   cd pi-zyloo-provider

2. Set your Zyloo API key:

   # Recommended: add to auth.json
   # See Authentication section below
   
   # Or set as environment variable
   export ZYLOO_API_KEY=your-api-key-here

3. Run pi with the extension:

   pi -e /path/to/pi-zyloo-provider

Option 3: Via npm

Install the package from npm:

npm install pi-zyloo-provider
# or
pnpm add pi-zyloo-provider

Then load it by path (npm installs into node_modules):

pi -e ./node_modules/pi-zyloo-provider

Authentication

The Zyloo API key can be configured in multiple ways (resolved in this order):

1. auth.json (recommended) — Add to ~/.pi/agent/auth.json:

   { "zyloo": { "type": "api_key", "key": "your-api-key" } }

   The key field supports literal values, env var names, and shell commands (prefix with !). See pi's auth file docs for details.
2. Runtime override — Use the --api-key CLI flag
3. Environment variable — Set ZYLOO_API_KEY

Get your API key from zyloo.io.

Environment Variables

| Variable | Required | Description | |----------|----------|-------------| | ZYLOO_API_KEY | No | Your Zyloo API key (fallback if not in auth.json) |

Configuration

Add to your pi configuration for automatic loading:

{
  "extensions": [
    "/path/to/pi-zyloo-provider"
  ]
}

Usage

Once loaded, select a model with:

/model zyloo zyloo/claude-opus-4-8

Or use /models to browse all available Zyloo models.

Reasoning Effort

For reasoning models, control thinking depth:

/reasoning high

Values: none, low, medium, high (and xhigh/max where the model exposes it via thinkingLevelMap).

API Compatibility

Zyloo provides an OpenAI-compatible Chat Completions proxy. Reasoning format and role handling are configured per model family in patch.json:

| Family | Models | thinkingFormat | Reasoning Effort | Developer Role | |--------|-------|-------------------|------------------|---------------| | Anthropic Claude | claude-* | openai | ✅ | ✅ | | OpenAI GPT (reasoning) | gpt-5.* | openai | ✅ | ✅ | | OpenAI GPT (chat) | gpt-4.1, gpt-4o | — | — | ✅ | | Google Gemini | gemini-* | openai | ✅ | ✅ | | xAI Grok | grok-* | — (inherent) | ❌ | ❌ (system) | | Moonshot Kimi | kimi-* | deepseek | ❌ | ❌ (system) | | Zhipu GLM | glm-5.2 | openai | ✅ | ❌ (system) | | Zhipu GLM | glm-4.7, glm-5, glm-5.1 | openai | ❌ | ❌ (system) | | DeepSeek | deepseek-* | deepseek | ✅ (high/max) | ❌ (system) | | Qwen | qwen3.7-max | openai | ✅ | ✅ | | Qwen | qwen3-max (chat) | — | — | ✅ | | MiniMax | minimax-* | openai | ✅ | ✅ |

Key notes:

| Aspect | Behavior | |--------|----------| | API | openai-completions (Chat Completions) | | Base URL | https://api.zyloo.io/v1 | | Max tokens field | Both max_tokens and max_completion_tokens accepted | | Reasoning output | reasoning_content in responses | | store parameter | Not sent (supportsStore: false) |

The Zyloo /v1/models endpoint only returns model IDs and owned_by — no pricing, context lengths, or reasoning flags. patch.json provides these based on each upstream provider's published specifications.

Updating Models

Run the update script to fetch the latest models from Zyloo's API:

export ZYLOO_API_KEY=your-api-key
node scripts/update-models.js

This will:

1. Fetch models from https://api.zyloo.io/v1/models
2. Preserve pricing and compat from existing models.json
3. Apply overrides from patch.json
4. Update models.json and the README model table

A GitHub Actions workflow can run this daily and open a PR if models have changed.

License

MIT