Is a local model as good as Claude?

The app says plainly that a local 8B model is not Claude’s equal on code; its advantages are that it is free, offline and nothing leaves the machine.

Why can’t I point the local engine to another computer?

The connector only accepts loopback addresses, so it cannot be used as an outgoing proxy from the interface.

AI model “Unreachable”, “Not installed on this machine” or “Groq refused it”

The model picker of the Senzoukria assistant tests the engine you choose. “Unreachable” means it did not answer, without saying whether the key, the network or the engine is at fault; the raw answer is shown under it. Local engines must run on this machine, and Groq keys are tested before use.

Senzoukria · Troubleshooting · Updated September 2026

At a glance

Unreachable: “Measured: the model did not answer. On its own that says nothing about which of the key, the network or the engine is at fault…”
No local model: “Not installed on this machine”
Groq test: “Key accepted · {n} models served right now” / “Groq refused it: {error}”
Local address: Loopback only (Ollama 11434, LM Studio 1234)

What you see

In the assistant’s model picker, the engine state is “Testing…”, “Available” (“Answered in {ms} ms.”) or “Unreachable” with “Measured: the model did not answer. On its own that says nothing about which of the key, the network or the engine is at fault — what came back is below.” and the raw error. With the engine up but no model, “Not installed on this machine” appears above “Models worth pulling”. For Groq, “Test the key” answers “Key accepted · {n} models served right now” or “Groq refused it: {error}”.

Why it happens

A local engine runs on your own CPU or GPU. The app only talks to engines on this machine: addresses must be localhost or a loopback IP, and must stop at the port. Ollama listens on port 11434 by default, LM Studio on 1234. If the engine program is not running, the test is “Unreachable”. If it runs without any model pulled, there is nothing to answer with. Groq is a hosted service on your own key, rate-limited per minute on its free tier.

How to fix it

Start Ollama (or LM Studio) on this machine, then click “Test again”.
When the engine answers but has no model, click Install next to a suggested model, or paste the shown command into a terminal.
Pick the model itself (for instance qwen3:8b), not Ollama: the picker notes that Ollama is the program, not a model.
For Groq, paste a key from console.groq.com, save it and click “Test the key”.

What it is not

Not a proof that your key is wrong: Unreachable does not say which part failed.
Not a remote server: the local connector refuses any address that is not on this machine.

When to contact support

If the problem persists, contact support through Menu → Report a problem, or on the Senzoukria Discord. Copy the raw error shown under the state, the engine and the model name.

In the same section

This page in other languages

Frequently asked questions

Is a local model as good as Claude?: The app says plainly that a local 8B model is not Claude’s equal on code; its advantages are that it is free, offline and nothing leaves the machine.
Why can’t I point the local engine to another computer?: The connector only accepts loopback addresses, so it cannot be used as an outgoing proxy from the interface.

Keep reading

Strategy researchTest a trading strategy, from observation to challenge rulesDefine an observation, question it with the copilot, write a rule and explore costs, drawdown and fictional challenge limits in an interactive example.Quant · Start hereQuant trading for beginners: from idea to backtestDefine a testable rule, choose suitable data and evaluate later periods after costs.Order flow · HubOrder flow trading: the complete guideThe hub. What order flow is, what it shows — aggression, delta, imbalance, absorption, liquidity — and the tools that read it, each linked to a full guide.