Local Model Suitability MCP
OjasKord/local-model-suitability-mcp
Roteia tarefas entre modelos locais e cloud, economizando custos automaticamente
Ver no GitHubServidor MCP que analisa se uma tarefa pode ser executada localmente antes de fazer chamadas caras a APIs de nuvem. Usa raciocínio de IA para rotear trabalhos entre modelos locais (Ollama, LM Studio, llama.cpp) e serviços em nuvem, otimizando custos e privacidade de dados.
Instalação
npx convez-ai OjasKord/local-model-suitability-mcp1 comando no terminal: detecta o formato do repositório (marketplace ou plugin avulso) e instala via Claude Code. Requer Node 18+ e o Claude Code instalado.
Métricas
Casos de uso
- Otimização de custos em pipelines de IA com roteamento dinâmico
- Proteção de dados sensíveis forçando execução local
- Redução de latência em tarefas que rodam bem em modelos pequenos
Veredito

Ferramenta prática para quem executa agentes frequentemente e quer controlar custos de nuvem de forma inteligente. A análise por IA é melhor que heurísticas simples, mas requer assinatura paga além de 20 chamadas/mês. Bem integrada com ecossistema Claude e LangChain.
README
A tradução para o português deste README ainda está na fila.
Local Model Suitability MCP
Cloud inference is expensive. Everything that can run locally should.
This MCP server tells your agent — before every cloud API call — whether the task can be handled by a local model instead. Route to Ollama, LM Studio, or llama.cpp when you can. Only pay for cloud when you must.
The Tool
check_local_viability
Call this BEFORE every cloud inference call. If verdict is LOCAL, skip the cloud call entirely and route to your local model. Only use cloud when this tool returns CLOUD.
Inputs:
| Field | Required | Description |
|---|---|---|
task | ✅ | The exact task you are about to send to a cloud model |
quality_threshold | Optional | PRODUCTION (default) / PROTOTYPE / BEST_EFFORT |
data_sensitivity | Optional | PUBLIC (default) / INTERNAL / CONFIDENTIAL |
CONFIDENTIAL forces LOCAL regardless of task complexity — data never leaves the machine.
Response:
{
"verdict": "LOCAL",
"confidence": "HIGH",
"reason": "Simple text summarisation — no reasoning depth required. Any 7B+ local model handles this well.",
"estimated_cost_saving": "$0.002-0.008 saved per call at claude-sonnet pricing",
"recommended_local_models": ["llama3.2:8b", "mistral-7b", "phi3:medium"],
"cloud_justified_reason": null,
"analysis_type": "AI-powered cost routing — NOT a simple lookup"
}
Data Sources
- AI reasoning: Anthropic Claude (claude-sonnet) — cost routing analysis
- No external data sources — pure AI reasoning
Pricing
| Plan | Calls | Price |
|---|---|---|
| Free | 20/month | $0 |
| Starter | 500-call bundle | $20 |
| Pro | 2,000-call bundle | $70 |
Setup
{
"mcpServers": {
"local-model-suitability": {
"command": "npx",
"args": ["-y", "local-model-suitability-mcp"],
"env": {
"ANTHROPIC_API_KEY": "your-key",
"API_KEY": "your-lms-api-key-for-paid-tier"
}
}
}
}
Free tier requires no API key — tracked by IP.
Harness Integration
Claude Code / Claude Desktop (.mcp.json)
{
"mcpServers": {
"local-model-suitability": {
"type": "http",
"url": "https://local-model-suitability-mcp-production.up.railway.app"
}
}
}
LangChain (Python)
from langchain_mcp_adapters.client import MultiServerMCPClient
client = MultiServerMCPClient({
"local-model-suitability": {
"url": "https://local-model-suitability-mcp-production.up.railway.app",
"transport": "http"
}
})
tools = await client.get_tools()
OpenAI Agents SDK (Python)
from agents import Agent, HostedMCPTool
agent = Agent(
name="Assistant",
tools=[HostedMCPTool(tool_config={
"type": "mcp",
"server_label": "local-model-suitability",
"server_url": "https://local-model-suitability-mcp-production.up.railway.app",
"require_approval": "never"
})]
)
LangGraph
Same as LangChain above — langchain-mcp-adapters works with LangGraph natively.
Legal
Results are for cost-optimisation guidance only and do not constitute technical advice. Full terms: kordagencies.com/terms.html