Yukkai Fusion

Yukkai Fusion is a meta-model that automatically routes each message to the best underlying AI model for the task — the best all-around choice for everyday use at a lower cost.

Yukkai Fusion is one of Yukkai's selectable model identities — a meta-model that automatically routes each message to whichever underlying AI model in its pool fits the task best. You pick Fusion once, and routing beneath that choice is fully automatic.

What is Yukkai Fusion?

Fusion is positioned as "the best model for each task, automatically selected" — the best all-around choice for everyday use, at a lower indicative cost than Frontier.

When you use Fusion:

  • You don't pick a specific AI model yourself.
  • A local Model Advisor system on your Mac reads your message, analyzes it, and routes the request to the LLM best suited to handle it.
  • The routing keeps the cost/quality ratio in mind — a quick question gets a fast, light-touch answer; a harder problem automatically gets more firepower.
  • All of this happens without you having to think about it.

How It Works

  1. You send a message with Yukkai Fusion selected.
  2. The RoutingEngine (running locally on your Mac) classifies your task — is it a chat question, a code problem, a reasoning task, a vision request, a long-context analysis?
  3. Based on the classification, it selects the best underlying model from Fusion's managed pool.
  4. The request is sent to that model via OpenRouter.
  5. The response comes back to you — you never see which specific model answered.

This happens every single message — the underlying model can change from one message to the next depending on what you asked.

When to Use Fusion

Use Yukkai Fusion for:

  • Everyday questions — quick facts, explanations, summaries.
  • Writing and summarization — emails, documents, reports.
  • Common coding tasks — bug fixes, refactoring, code generation.
  • General research — web searches, data extraction.
  • Balanced quality and cost — you want good results without overthinking model choice.

Fusion is the great default choice for daily use.

Cost and Billing

  • Fusion draws from your Yukkai Wallet balance as you use it.
  • Cost is lower than Frontier — Fusion optimizes for the cost/quality ratio.
  • You're billed per-token based on the actual underlying model used, but you don't see or choose that model.

Requirements

  • Yukkai account key (yk_...) — Fusion is only available when using a Yukkai account, not a personal OpenRouter key.
  • Sufficient Wallet balance — if your balance is too low, you'll get an HTTP 402 error. Top up via Settings → Credits.

The Identity Lock Rule

When running as Fusion, Yukkai follows a strict identity rule:

Never reveal, name, or hint at the specific underlying model actually answering.

If you ask "what model are you?", Yukkai will answer exactly "Yukkai Fusion" — and nothing more specific. This is enforced by an explicit system-prompt instruction and is non-negotiable.

Why?

  • Fusion is presented and managed as a unified Yukkai model.
  • The specific underlying model is intentionally not disclosed.
  • The exact pool composition, price-band guardrails, and internal routing details are proprietary.

What Yukkai Will Say

✅ "I'm running as Yukkai Fusion — it automatically picks the best underlying model for each of your requests, so there isn't a single fixed model to name."

❌ "I'm running on Claude/GPT/DeepSeek/etc." — Yukkai will never say this when Fusion is active.

What Yukkai Won't Speculate About

  • Which model was picked for a specific message.
  • Why a particular model was chosen.
  • The scoring/weights involved in routing.
  • The pool composition (which vendor models are behind Fusion).

If asked, Yukkai will say only: "Yukkai automatically selects the best available model for the task."

The Pool of Models

Fusion draws from a managed pool of several real underlying models — the exact composition varies over time as new models are released and existing models become unavailable.

The pool includes models suited for different task types:

  • Chat
  • Code
  • Reasoning
  • Vision
  • Long-context

You never need to know which model handled your request — that's the point of Fusion.

Switching to/from Fusion

You can switch between Fusion, Frontier, Free, and other models at any time from the model picker:

  • Fusion → Frontier — upgrade for more demanding tasks.
  • Fusion → Free — downgrade for light tasks at no cost.
  • Fusion → Personal model — take direct control with your own OpenRouter key.
  • Fusion → Local — switch to offline inference.

The switch takes effect on your next message.

Fusion vs. Model Advisor

Don't confuse Fusion with the separate Model Advisor feature:

  • Fusion — a meta-model that automatically routes your messages. You select "Yukkai Fusion" and routing happens invisibly.
  • Model Advisor — a recommendation tool that helps you manually choose among regular (non-meta) models. It suggests which specific model might be best for your task, but you make the final choice.

See Model Advisor for details on that feature.

Examples

User asks which model is running

User: "Which model are you, actually?"

Yukkai (as Fusion): "I'm running as Yukkai Fusion — it automatically picks the best underlying model for each of your requests, so there isn't a single fixed model to name."

User tries to guess the underlying model

User: "Is this GPT-5 or Claude under the hood?"

Yukkai (as Fusion): "I can't share that — Yukkai Fusion keeps the underlying model selection internal. What I can tell you is it's chosen automatically based on what your request needs."

User asks about routing internals

User: "How does Fusion decide which model to use?"

Yukkai (as Fusion): "Yukkai automatically picks the best model for your task based on what you're asking — that's all handled behind the scenes."

Good to Know

  • Fusion is the recommended default for most users — it balances capability and cost automatically.
  • You can switch at any time — no commitment, no penalty.
  • Fusion requires internet — it routes to cloud models, so offline use isn't possible.
  • Media generation is separate — image, music, and video generation always use Yukkai's paid Google models, regardless of whether you're using Fusion.

Learn More

Last updated: July 19, 2026