Skip to main content

​
AI Router

​
Overview

SmolAgents is a lightweight Python agent framework by Hugging Face. Using OpenAIServerModel with a custom api_base, you can route all LLM calls through Orq.ai’s AI Router for access to 300+ models, cost tracking, and reliability features.

​
Prerequisites

  • An Orq.ai account and API Key
  • Python 3.9 or higher
To setup your API key, see API keys & Endpoints.

​
Installation

pip install smolagents

​
Configuration

Python
import os
from smolagents import OpenAIServerModel

model = OpenAIServerModel(
    model_id="openai/gpt-4o",
    api_base="https://api.orq.ai/v2/router",
    api_key=os.environ["ORQ_API_KEY"],
)
api_base: https://api.orq.ai/v2/router

​
Basic Agent Example

Python
import os
from smolagents import CodeAgent, OpenAIServerModel

model = OpenAIServerModel(
    model_id="openai/gpt-4o-mini",
    api_base="https://api.orq.ai/v2/router",
    api_key=os.environ["ORQ_API_KEY"],
)

agent = CodeAgent(tools=[], model=model)
agent.run("What is the capital of France?")

​
Model Selection

With Orq.ai, you can use any supported model from 20+ providers:
Python
import os
from smolagents import OpenAIServerModel

# Use Claude
claude_model = OpenAIServerModel(
    model_id="anthropic/claude-sonnet-4-5-20250929",
    api_base="https://api.orq.ai/v2/router",
    api_key=os.environ["ORQ_API_KEY"],
)

# Use Gemini
gemini_model = OpenAIServerModel(
    model_id="google/gemini-2.5-flash",
    api_base="https://api.orq.ai/v2/router",
    api_key=os.environ["ORQ_API_KEY"],
)

# Use Groq
groq_model = OpenAIServerModel(
    model_id="groq/llama-3.3-70b-versatile",
    api_base="https://api.orq.ai/v2/router",
    api_key=os.environ["ORQ_API_KEY"],
)

​
Observability

​
Overview

SmolAgents uses the openinference-instrumentation-smolagents library to automatically instrument agent runs and export traces via OpenTelemetry.

​
Prerequisites

  • An Orq.ai account and API Key
  • Python 3.9 or higher

​
Installation

pip install smolagents \
    opentelemetry-api \
    opentelemetry-sdk \
    "opentelemetry-exporter-otlp-proto-http" \
    openinference-instrumentation-smolagents

​
Configuring Orq.ai Observability

Set up the tracer provider and instrument SmolAgents before creating any agents: 
import os
from opentelemetry.sdk.trace import TracerProvider
from opentelemetry.sdk.trace.export import BatchSpanProcessor
from opentelemetry.exporter.otlp.proto.http.trace_exporter import OTLPSpanExporter
from openinference.instrumentation.smolagents import SmolagentsInstrumentor

exporter = OTLPSpanExporter(
    endpoint="https://api.orq.ai/v2/otel/v1/traces",
    headers={"Authorization": f"Bearer {os.environ['ORQ_API_KEY']}"},
)
tracer_provider = TracerProvider()
tracer_provider.add_span_processor(BatchSpanProcessor(exporter))

SmolagentsInstrumentor().instrument(tracer_provider=tracer_provider)
SmolagentsInstrumentor must be called before any agent is instantiated. Once instrumented, all CodeAgent and ToolCallingAgent runs are automatically traced.

​
Basic Example

import os
from opentelemetry.sdk.trace import TracerProvider
from opentelemetry.sdk.trace.export import BatchSpanProcessor
from opentelemetry.exporter.otlp.proto.http.trace_exporter import OTLPSpanExporter
from openinference.instrumentation.smolagents import SmolagentsInstrumentor

exporter = OTLPSpanExporter(
    endpoint="https://api.orq.ai/v2/otel/v1/traces",
    headers={"Authorization": f"Bearer {os.environ['ORQ_API_KEY']}"},
)
tracer_provider = TracerProvider()
tracer_provider.add_span_processor(BatchSpanProcessor(exporter))
SmolagentsInstrumentor().instrument(tracer_provider=tracer_provider)

from smolagents import CodeAgent, OpenAIServerModel

agent = CodeAgent(
    tools=[],
    model=OpenAIServerModel(
        model_id="openai/gpt-4o-mini",
        api_base="https://api.orq.ai/v2/router",
        api_key=os.environ["ORQ_API_KEY"],
    ),
)
agent.run("What is the capital of France?")