A
AMIYO
THINK BEYOND. BUILT LOCAL.
New conversation
A
AMIYO
ONE AI INTELLIGENCE
AMIYO ยท INTELLIGENCE CORE

Hello!

Ask anything โ€” AMIYO reads your intent, routes to the right capability, and gets to work.

ContextGeneral
Voice Ready
โšก Auto (Best Model) ยท Intelligent Auto-Routing โ€ข Zero API keys ยท Enter to send ยท Shift + Enter for a new line
Search AMIYO
๐Ÿค– AI Model Selector โ€” Zero API Key & Open Models
Choose between Auto Routing, Built-in Offline Engine, Local Ollama, LM Studio, Pollinations, Kilo, HuggingFace & WebLLM (100% Zero-Auth)
โšก
Zero-Auth AI Ecosystem (0 API Keys Required) AUTO-CONNECTING...
Auto-detects and connects 16 Zero-Auth engines: Local Ollama, LM Studio, llama.cpp, KoboldCpp, Jan, LocalAI, WebLLM + Cloud Pollinations, Puter.js, Kilo, DuckDuckGo, Airforce, Blackbox, OVHcloud, LLM7, HuggingFace
๐ŸŒ
WebLLM In-Browser Engine CHECKING WEBGPU...
Runs 100% on client GPU inside browser tab โ€ข Zero server โ€ข Cached in IndexedDB
Select a model and click Load Model
๐Ÿฆ™
Ollama Local Engine http://localhost:11434
Checking connection...
Server offline. Start in 1-click or run: ollama run llama3.2:1b
๐Ÿค—
HuggingFace Open Inference Serverless Multi-Model Router
Zero-Auth Active
โšก Auto Mode: Dynamically routes coding to Qwen Coder, math/logic to DeepSeek R1, writing to Mistral, and science/general to Llama 70B.
๐Ÿ–ฅ๏ธ
LM Studio & Local GGUF Engine http://localhost:1234 โ€ข OpenAI-Compatible API
Checking localhost:1234...
Detecting local GGUF engine...
Auto-Failover Active: If LM Studio is not running on :1234, queries are instantly served by your active local GGUF engine.
๐ŸŒธ
Pollinations.ai & Free Cloud AI https://text.pollinations.ai โ€ข Zero API Key & Zero Sign-up
Zero-Auth Active
100% Free Public Inference: Zero auth required. Direct CORS support in browser.
โšก
Kilo Free Gateway High-Capacity Free Open Model Gateway
Free Pool Active
High-speed serverless pools: Laguna S, Nemotron Super & GPT-OSS with zero sign-up.
โš™๏ธ
llama.cpp, vLLM & KoboldCpp :8080 โ€ข :8000 โ€ข :5001 โ€ข :1337 โ€ข Raw C++ Engine
Ollama Auto-Failover Ready
Auto-Failover Active: Routes to llama-server (:8080), vLLM (:8000), KoboldCpp (:5001), Jan (:1337) or Ollama (:11434).
โšก
Puter.js Frontier Cloud AI (100% Free) Claude 3.5 Sonnet, GPT-4o, DeepSeek R1 โ€ข Zero API Key & Zero Sign-up
Frontier Pool Ready
In-Browser Client Execution: Runs directly in your browser session with 0 API tokens required.
๐Ÿฆ†
DuckDuckGo Anonymous Privacy AI Zero Sign-Up โ€ข Zero Token Tracking โ€ข 100% Anonymous
Anonymous Active
Privacy-First AI: Completely private requests with zero authentication or user profile linkage.
โœˆ๏ธ
Airforce AI & Blackbox Coding Engine DeepSeek R1, Qwen Coder, Blackbox Code โ€ข Zero API Key
Router Active
Free Open Router: Multi-provider failover across DeepSeek, Qwen, Llama & Blackbox.
๐Ÿฐ
LocalAI, Tabby & Text-Generation-WebUI :5000 โ€ข :2242 โ€ข :7860 โ€ข Open-Source Server Hub
Localhost Ready
Scans local AI servers on ports 5000 (Oobabooga), 2242 (Aphrodite), 7860 (Gradio), and 8080 (LocalAI).
Automatic localhost discovery with immediate failover to Ollama (:11434).
๐Ÿ”‘
Keyed & Custom AI Providers APInex, Groq, OpenRouter, Cerebras, Together AI, DeepInfra, Mistral, OpenAI, Ashna, 1min, Experiential
๐Ÿ”’ Keys Stored in Browser LocalStorage Only
๐Ÿค—
HuggingFace Access Token
Configure your Hugging Face User Access Token for Serverless Inference
๐Ÿ“Œ How it works: โ€ข Zero-Auth Mode (Default): If left blank, AMIYO automatically uses the Zero-Auth open tier so you can use HuggingFace open-weight intelligence directly with zero API keys.
โ€ข With Access Token: Queries are routed with higher priority and throughput to Hugging Face Inference Providers, unlocking gated frontier weights (Llama 3.3 70B, DeepSeek R1, Mistral Large).
โ€ข Get a free token: Visit huggingface.co/settings/tokens and create a token with Inference Providers or Read permissions.
โ€ข Privacy: Your token is saved only in your local browser (localStorage) and sent directly to Hugging Face.
๐Ÿค—
Hugging Face 3M+ Model Hub Explorer 3,000,000+ MODELS Zero-Auth Ready
Search, live-test, and deploy any open-weights model directly into AMIYO inference router
๐Ÿ”
Fast Picks: ๐Ÿ‹ DeepSeek-R1 ๐Ÿ’ป Qwen 2.5 Coder ๐Ÿฆ™ Llama 3.3 ๐ŸŒช๏ธ Mistral Small ๐Ÿ”ฌ Phi-4 ๐Ÿ’ก QwQ 32B ๐Ÿš€ Codestral ๐Ÿ’Ž Gemma 2
Deploy Any Model Directly from 3M+ Repos:
Browsing top Hugging Face models...
๐Ÿค–
Multi-Model Joint Agent Studio
Collaboratively link reasoning, coding, web search & image generation models into frontier autonomous teams
๐Ÿš€ Full-Stack AI Architect
โœจ Frontier Joint Teams orchestrate multiple specialized open weights models in unison. Each model handles the task it excels at โ€” from mathematical deduction to full-stack code, real-time web grounding, and high-fidelity image rendering.
โญ Your Custom Joint Agent Teams
Tool Manager
Add files
ID
AMIYO Universal Canonical Identity Fabric Explorer 1,032 Canonical Objects ยท 324 Legacy Mappings ยท 44 Relationship Edges ยท Zero Collisions
Select an object from the list to inspect its Universal Header, Aliases, Typo Variants, and Relationship Graph.
๐Ÿ›ก๏ธ Fabric Health: 100% HEALTHY ๐Ÿ” Idempotency: Stable Checkpoint โšก Collision Guard: Zero Overlaps
๐ŸŽ™๏ธ Voice Assistant & Multi-Country Accent Studio
๐Ÿ‡ฎ๐Ÿ‡ณ
Priya ๐Ÿ‘ฉ Female โœจ Ultra-Natural Neural ยท India (en-IN)
Warm, natural Indian English female accent โ€” calm, neutral human cadence for lessons and problem solving.
Voice Cadence:
1.0x
1.0x
๐Ÿ—ฃ๏ธ Explore All Voices 24 Available (12 Female ยท 12 Male across 10 Countries)
Gender:
Profile
Explore
โšก Live Website Preview LIVE WYSIWYG
โŒ˜ ESC to close
Letter Templates