Strands Agents

Strands Agents is a simple-to-use agent framework built by AWS. Portkey enhances Strands Agents with production-grade observability, reliability, and multi-provider support—all through a single integration that requires no changes to your existing agent logic.

What you get with this integration:

Complete observability of every agent step, tool use, and LLM interaction
Built-in reliability with automatic fallbacks, retries, and load balancing
200+ LLMs accessible through the same OpenAI-compatible interface
Production monitoring with traces, logs, and real-time metrics
Zero code changes to your existing Strands agent implementations

Strands Agents Documentation

Learn more about Strands Agents’ core concepts and features

Quick Start

Install Dependencies

pip install -U strands-agents strands-agents-tools openai portkey-ai

Replace Your Model Initialization

Instead of initializing your OpenAI model directly:

# Before: Direct OpenAI
from strands.models.openai import OpenAIModel

model = OpenAIModel(
    client_args={"api_key": "sk-..."},
    model_id="gpt-4o",
    params={"temperature": 0.7}
)

Initialize it through Portkey’s gateway:

# After: Through Portkey
from strands.models.openai import OpenAIModel
from portkey_ai import PORTKEY_GATEWAY_URL

model = OpenAIModel(
    client_args={
        "api_key": "YOUR_PORTKEY_API_KEY",  # Your Portkey API key
        "base_url": PORTKEY_GATEWAY_URL     # Routes through Portkey
    },
    model_id="gpt-4o",
    params={"temperature": 0.7}
)

Use Your Agent Normally

from strands import Agent
from strands_tools import calculator

agent = Agent(model=model, tools=[calculator])
response = agent("What is 2+2?")
print(response)

Your agent works exactly the same way, but now all interactions are automatically logged, traced, and monitored in your Portkey dashboard.

How the Integration Works

The integration leverages Strands’ flexible client_args parameter, which passes any arguments directly to the OpenAI client constructor. By setting base_url to Portkey’s gateway, all requests route through Portkey while maintaining full compatibility with the OpenAI API.

# This is what happens under the hood in Strands:
client_args = client_args or {}
self.client = openai.OpenAI(**client_args)  # Your Portkey config gets passed here

This means you get all of Portkey’s features without any changes to your agent logic, tool usage, or response handling.

Setting Up Portkey

Before using the integration, you need to configure your AI providers and create a Portkey API key.

Add Your AI Provider Keys

Go to Virtual Keys in the Portkey dashboard and add your actual AI provider keys (OpenAI, Anthropic, etc.). Each provider key gets a virtual key ID that you’ll reference in configs.

Create a Configuration

Go to Configs to define how requests should be routed. A basic config looks like:

{
 "virtual_key": "openai-key-abc123"
}

For production setups, you can add fallbacks, load balancing, and conditional routing here.

Generate Your Portkey API Key

Go to API Keys to create a new API key. Attach your config as the default routing config, and you’ll get an API key that routes to your configured providers.

Complete Integration Example

Here’s a full example showing how to set up a Strands agent with Portkey integration:

from strands import Agent
from strands.models.openai import OpenAIModel
from strands_tools import calculator, web_search
from portkey_ai import PORTKEY_GATEWAY_URL

# Initialize model through Portkey
model = OpenAIModel(
    client_args={
        "api_key": "YOUR_PORTKEY_API_KEY",
        "base_url": PORTKEY_GATEWAY_URL
    },
    model_id="gpt-4o",
    params={
        "max_tokens": 1000,
        "temperature": 0.7,
    }
)

# Create agent with tools (unchanged from standard Strands usage)
agent = Agent(
    model=model, 
    tools=[calculator, web_search]
)

# Use the agent (unchanged from standard Strands usage)
response = agent("Calculate the compound interest on $10,000 at 5% for 10 years, then search for current inflation rates")
print(response)

The agent will automatically use both tools as needed, and every step will be logged in your Portkey dashboard with full request/response details, timing, and token usage.

Production Features

1. Enhanced Observability

Portkey provides comprehensive visibility into your agent’s behavior without requiring any code changes.

Track the complete execution flow of your agents with hierarchical traces that show:

LLM calls: Every request to language models with full payloads
Tool invocations: Which tools were called, with what parameters, and their responses
Decision points: How the agent chose between different tools or approaches
Performance metrics: Latency, token usage, and cost for each step

from strands import Agent
from strands.models.openai import OpenAIModel
from strands_tools import calculator
from portkey_ai import PORTKEY_GATEWAY_URL, createHeaders

model = OpenAIModel(
    client_args={
        "api_key": "YOUR_PORTKEY_API_KEY",
        "base_url": PORTKEY_GATEWAY_URL,
        # Add trace ID to group related requests
        "default_headers": createHeaders(trace_id="user_session_123")
    },
    model_id="gpt-4o",
    params={"temperature": 0.7}
)

agent = Agent(model=model, tools=[calculator])
response = agent("What's 15% of 2,847?")

All requests from this agent will be grouped under the same trace, making it easy to analyze the complete interaction flow.

Track the complete execution flow of your agents with hierarchical traces that show:

LLM calls: Every request to language models with full payloads
Tool invocations: Which tools were called, with what parameters, and their responses
Decision points: How the agent chose between different tools or approaches
Performance metrics: Latency, token usage, and cost for each step

from strands import Agent
from strands.models.openai import OpenAIModel
from strands_tools import calculator
from portkey_ai import PORTKEY_GATEWAY_URL, createHeaders

model = OpenAIModel(
    client_args={
        "api_key": "YOUR_PORTKEY_API_KEY",
        "base_url": PORTKEY_GATEWAY_URL,
        # Add trace ID to group related requests
        "default_headers": createHeaders(trace_id="user_session_123")
    },
    model_id="gpt-4o",
    params={"temperature": 0.7}
)

agent = Agent(model=model, tools=[calculator])
response = agent("What's 15% of 2,847?")

All requests from this agent will be grouped under the same trace, making it easy to analyze the complete interaction flow.

Add business context to your agent runs for better filtering and analysis:

from portkey_ai import createHeaders

model = OpenAIModel(
    client_args={
        "api_key": "YOUR_PORTKEY_API_KEY",
        "base_url": PORTKEY_GATEWAY_URL,
        "default_headers": createHeaders(
            trace_id="customer_support_bot",
            metadata={
                "agent_type": "customer_support",
                "user_tier": "premium",
                "session_id": "sess_789",
                "department": "billing"
            }
        )
    },
    model_id="gpt-4o",
    params={"temperature": 0.3}  # Lower temperature for support tasks
)

This metadata appears in your Portkey dashboard, allowing you to filter logs and analyze performance by user type, session, or any custom dimension.

Monitor your agents in production with built-in dashboards that track:

Success rates: Percentage of successful agent completions
Average latency: Response times across different agent types
Token usage: Track consumption and costs across models
Error patterns: Common failure modes and their frequency

All metrics can be segmented by the metadata you provide, giving you insights like “premium user agents have 15% higher success rates” or “billing department queries take 2x longer on average.”

2. Reliability & Fallbacks

When running agents in production, things can go wrong - API rate limits, network issues, or provider outages. Portkey’s reliability features ensure your agents keep running smoothly even when problems occur.

It’s simple to enable fallback in your Strands Agents by using a Portkey Config that you can attach at runtime or directly to your Portkey API key. Here’s an example of attaching a Config at runtime:

Configure multiple providers so your agents keep working even when one provider fails:

from portkey_ai import createHeaders

model = OpenAIModel(
    client_args={
        "api_key": "YOUR_PORTKEY_API_KEY", 
        "base_url": PORTKEY_GATEWAY_URL,
        "default_headers": createHeaders(
            config={
                "strategy": {
                    "mode": "fallback",
                    "on_status_codes": [429, 503, 502]  # Rate limits and server errors
                },
                "targets": [
                    { "virtual_key": "openai-key-primary" },   # Try OpenAI first
                    { "virtual_key": "anthropic-key-backup" }  # Fall back to Claude
                ]
            }
        )
    },
    model_id="gpt-4o",  # Will map to equivalent models on each provider
    params={"temperature": 0.7}
)

If OpenAI returns a rate limit error (429), Portkey automatically retries the request with Anthropic’s Claude, using default model mappings.

Configure multiple providers so your agents keep working even when one provider fails:

from portkey_ai import createHeaders

model = OpenAIModel(
    client_args={
        "api_key": "YOUR_PORTKEY_API_KEY", 
        "base_url": PORTKEY_GATEWAY_URL,
        "default_headers": createHeaders(
            config={
                "strategy": {
                    "mode": "fallback",
                    "on_status_codes": [429, 503, 502]  # Rate limits and server errors
                },
                "targets": [
                    { "virtual_key": "openai-key-primary" },   # Try OpenAI first
                    { "virtual_key": "anthropic-key-backup" }  # Fall back to Claude
                ]
            }
        )
    },
    model_id="gpt-4o",  # Will map to equivalent models on each provider
    params={"temperature": 0.7}
)

If OpenAI returns a rate limit error (429), Portkey automatically retries the request with Anthropic’s Claude, using default model mappings.

Distribute requests across multiple API keys to stay within rate limits:

from portkey_ai import createHeaders

model = OpenAIModel(
    client_args={
        "api_key": "YOUR_PORTKEY_API_KEY",
        "base_url": PORTKEY_GATEWAY_URL, 
        "default_headers": createHeaders(
            config={
                "strategy": {"mode": "loadbalance"},
                "targets": [
                    { "virtual_key": "openai-key-1", "weight": 70 },
                    { "virtual_key": "openai-key-2", "weight": 30 }
                ]
            }
        )
    },
    model_id="gpt-4o",
    params={"temperature": 0.7}
)

Requests will be distributed 70/30 across your two OpenAI keys, helping you maximize throughput without hitting individual key limits.

Route requests to different providers/models based on custom logic (like metadata, input content, or user attributes) using Portkey’s Conditional Routing feature.

See the Conditional Routing documentation for full guidance and advanced examples.

3. LLM Interoperability

Access 1,600+ models through the same Strands interface by changing just the provider configuration:

from portkey_ai import createHeaders

# Use Claude instead of GPT-4
model = OpenAIModel(
    client_args={
        "api_key": "YOUR_PORTKEY_API_KEY",
        "base_url": PORTKEY_GATEWAY_URL,
        "default_headers": createHeaders(
            provider="anthropic",
            api_key="YOUR_ANTHROPIC_KEY"  # Can also use virtual keys
        )
    },
    model_id="claude-3-7-sonnet-latest",  # Claude model ID
    params={"max_tokens": 1000, "temperature": 0.7}
)

# Agent code remains exactly the same
agent = Agent(model=model, tools=[calculator])
response = agent("Explain quantum computing in simple terms")

from portkey_ai import createHeaders

# Use Claude instead of GPT-4
model = OpenAIModel(
    client_args={
        "api_key": "YOUR_PORTKEY_API_KEY",
        "base_url": PORTKEY_GATEWAY_URL,
        "default_headers": createHeaders(
            provider="anthropic",
            api_key="YOUR_ANTHROPIC_KEY"  # Can also use virtual keys
        )
    },
    model_id="claude-3-7-sonnet-latest",  # Claude model ID
    params={"max_tokens": 1000, "temperature": 0.7}
)

# Agent code remains exactly the same
agent = Agent(model=model, tools=[calculator])
response = agent("Explain quantum computing in simple terms")

# Create different model instances for different tasks
reasoning_model = OpenAIModel(
    client_args={
        "api_key": "YOUR_PORTKEY_API_KEY",
        "base_url": PORTKEY_GATEWAY_URL,
        "default_headers": createHeaders(virtual_key="openai-key")
    },
    model_id="gpt-4o",
    params={"temperature": 0.1}  # Low temperature for reasoning
)

creative_model = OpenAIModel(
    client_args={
        "api_key": "YOUR_PORTKEY_API_KEY", 
        "base_url": PORTKEY_GATEWAY_URL,
        "default_headers": createHeaders(virtual_key="gemini-key")
    },
    model_id="gemini-2.0-flash-exp",
    params={"temperature": 0.8}  # Higher temperature for creativity
)

# Use different models for different agent types
reasoning_agent = Agent(model=reasoning_model, tools=[calculator])
creative_agent = Agent(model=creative_model, tools=[])

Portkey provides access to LLMs from providers including:

OpenAI (GPT-4o, GPT-4 Turbo, etc.)
Anthropic (Claude 3.5 Sonnet, Claude 3 Opus, etc.)
Mistral AI (Mistral Large, Mistral Medium, etc.)
Google Vertex AI (Gemini 1.5 Pro, etc.)
Cohere (Command, Command-R, etc.)
AWS Bedrock (Claude, Titan, etc.)
Local/Private Models

Supported Providers

See the full list of LLM providers supported by Portkey

4. Guardrails for Safe Agents

Guardrails ensure your Strands agents operate safely and respond appropriately in all situations.

Why Use Guardrails?

Strands agents can experience various failure modes:

Generating harmful or inappropriate content
Leaking sensitive information like PII
Hallucinating incorrect information
Generating outputs in incorrect formats

Portkey’s guardrails can:

Detect and redact PII in both inputs and outputs
Filter harmful or inappropriate content
Validate response formats against schemas
Check for hallucinations against ground truth
Apply custom business logic and rules

Learn More About Guardrails

Explore Portkey’s guardrail features to enhance agent safety

Advanced Configuration

Configure different behavior for development, staging, and production:

import os
from portkey_ai import createHeaders

def create_model(environment="production"):
    if environment == "development":
        # Use faster, cheaper models for development
        headers = createHeaders(
            config={"targets": [{"virtual_key": "openai-dev-key"}]},
            metadata={"environment": "dev"}
        )
        model_id = "gpt-4o-mini"
        params = {"temperature": 0.5, "max_tokens": 500}
    
    elif environment == "production":
        # Use high-performance models with fallbacks for production
        headers = createHeaders(
            config={
                "strategy": {"mode": "fallback"},
                "targets": [
                    {"virtual_key": "openai-prod-key"},
                    {"virtual_key": "anthropic-prod-key"}
                ]
            },
            metadata={"environment": "prod"}
        )
        model_id = "gpt-4o"
        params = {"temperature": 0.7, "max_tokens": 2000}
    
    return OpenAIModel(
        client_args={
            "api_key": "YOUR_PORTKEY_API_KEY",
            "base_url": PORTKEY_GATEWAY_URL,
            "default_headers": headers
        },
        model_id=model_id,
        params=params
    )

# Use environment-specific configuration
model = create_model(os.getenv("APP_ENV", "production"))
agent = Agent(model=model, tools=[calculator])

Configure different behavior for development, staging, and production:

import os
from portkey_ai import createHeaders

def create_model(environment="production"):
    if environment == "development":
        # Use faster, cheaper models for development
        headers = createHeaders(
            config={"targets": [{"virtual_key": "openai-dev-key"}]},
            metadata={"environment": "dev"}
        )
        model_id = "gpt-4o-mini"
        params = {"temperature": 0.5, "max_tokens": 500}
    
    elif environment == "production":
        # Use high-performance models with fallbacks for production
        headers = createHeaders(
            config={
                "strategy": {"mode": "fallback"},
                "targets": [
                    {"virtual_key": "openai-prod-key"},
                    {"virtual_key": "anthropic-prod-key"}
                ]
            },
            metadata={"environment": "prod"}
        )
        model_id = "gpt-4o"
        params = {"temperature": 0.7, "max_tokens": 2000}
    
    return OpenAIModel(
        client_args={
            "api_key": "YOUR_PORTKEY_API_KEY",
            "base_url": PORTKEY_GATEWAY_URL,
            "default_headers": headers
        },
        model_id=model_id,
        params=params
    )

# Use environment-specific configuration
model = create_model(os.getenv("APP_ENV", "production"))
agent = Agent(model=model, tools=[calculator])

Override configuration for specific requests without changing the model:

from portkey_ai import createHeaders

model = OpenAIModel(
    client_args={
        "api_key": "YOUR_PORTKEY_API_KEY",
        "base_url": PORTKEY_GATEWAY_URL
    },
    model_id="gpt-4o",
    params={"temperature": 0.7}
)

# For high-priority requests, override to use faster/more reliable providers
agent = Agent(model=model, tools=[calculator])

# Normal request
response1 = agent("What's 2+2?")

# High-priority request with overrides (if your agent supports custom headers)
# This would require custom implementation in your agent wrapper
response2 = agent(
    "Critical calculation: What's the compound interest?",
    headers=createHeaders(
        config={"targets": [{"virtual_key": "premium-openai-key"}]},
        metadata={"priority": "high"}
    )
)

Enterprise Governance

If you are using Strands inside your organization, you need to consider several governance aspects:

Cost Management: Controlling and tracking AI spending across teams
Access Control: Managing which teams can use specific models
Usage Analytics: Understanding how AI is being used across the organization
Security & Compliance: Maintaining enterprise security standards
Reliability: Ensuring consistent service across all users

Centralized Key Management

Instead of distributing raw API keys to developers, use Portkey API keys that you can control centrally:

# Developers use Portkey API keys (not raw provider keys)
model = OpenAIModel(
    client_args={
        "api_key": "pk-team-frontend-xyz123",  # Team-specific Portkey key
        "base_url": PORTKEY_GATEWAY_URL
    },
    model_id="gpt-4o",
    params={"temperature": 0.7}
)

You can:

Rotate provider keys without updating any code
Set spending limits per team or API key
Control model access (which teams can use which models)
Monitor usage across all teams and projects
Revoke access instantly if needed

Usage Analytics & Budgets

Track and control AI spending across your organization:

Per-team budgets: Set monthly spending limits for different teams
Model usage analytics: See which teams are using which models most
Cost attribution: Understand costs by project, team, or user
Usage alerts: Get notified when teams approach their limits

All of this works automatically with your existing Strands agents—no code changes required.

Contact & Support

Enterprise SLAs & Support

Get dedicated SLA-backed support.

Portkey Community

Join our forums and Slack channel.

Resources

Strands Agents Docs

Book a Demo

Get personalized guidance on implementing this integration

Troubleshooting

Import Errors

Problem: ModuleNotFoundError when importing Portkey components

Solution: Ensure all dependencies are installed:

pip install -U strands-agents strands-agents-tools openai portkey-ai

Verify versions are compatible:

import strands
import portkey_ai
print(f"Strands: {strands.__version__}")
print(f"Portkey: {portkey_ai.__version__}")

Authentication Errors

Problem: AuthenticationError when making requests

Solution: Verify your Portkey API key and provider setup: Verify your Portkey API key and provider setup. Test your Portkey API key directly and check for common issues such as wrong API key format, misconfigured provider virtual keys, and missing config attachments.

# Test your Portkey API key directly
from portkey_ai import Portkey

portkey = Portkey(api_key="YOUR_PORTKEY_API_KEY")
response = portkey.chat.completions.create(
    messages=[{"role": "user", "content": "test"}],
    model="gpt-4o"
)
print(response)

Rate Limiting

Problem: Hitting rate limits despite having fallbacks configured

Solution: Check your fallback configuration:

# Ensure fallbacks are configured for rate limit status codes
headers = createHeaders(
    config={
        "strategy": {
            "mode": "fallback", 
            "on_status_codes": [429, 503, 502, 504]
        },
        "targets": [
            {"virtual_key": "primary-key"},
            {"virtual_key": "backup-key"}
        ]
    }
)

Also verify that your backup providers have sufficient quota.

Model Compatibility

Problem: Model not found or unsupported model errors

Solution: Check that your model ID is correct for the provider:

# OpenAI models
model_id = "gpt-4o"  # Correct
model_id = "gpt-4-turbo"  # Correct  

# Anthropic models  
model_id = "claude-3-5-sonnet-20241022"  # Correct
model_id = "claude-3-sonnet"  # Wrong format

# Use provider-specific model IDs, not generic names

Missing Traces/Logs

Problem: Not seeing traces or logs in Portkey dashboard

Solution: Verify your requests are going through Portkey:

# Check that base_url is set correctly
print(model.client.base_url)  # Should be https://api.portkey.ai/v1

# Add trace IDs for easier debugging
headers = createHeaders(
    trace_id="debug-session-123",
    metadata={"debug": "true"}
)

Also check the Logs section in your Portkey dashboard and filter by your metadata.

Frequently Asked Questions

How does Portkey enhance Strands Agents?

Can I use Portkey with existing Strands Agents applications?

Does Portkey work with all Strands Agents features?

Can I track usage across multiple agents in a workflow?

How do I filter logs and traces for specific agent runs?

Can I use my own API keys with Portkey?

Next Steps

Now that you have Portkey integrated with your Strands agents:

Monitor your agents in the Portkey dashboard to understand their behavior
Set up fallbacks for critical production agents using multiple providers
Add custom metadata to track different agent types or user segments
Configure budgets and alerts if you’re deploying multiple agents
Explore advanced routing to optimize for cost, latency, or quality

Portkey Dashboard

View your agent logs, traces, and analytics

Supported Models

See all 200+ models you can use with this integration

Enterprise Features

Explore governance, security, and compliance features

Phidata Bring Your own Agents

On this page

Quick Start
How the Integration Works
Setting Up Portkey
Complete Integration Example
Production Features
1. Enhanced Observability
2. Reliability & Fallbacks
3. LLM Interoperability
4. Guardrails for Safe Agents
Advanced Configuration
Enterprise Governance
Centralized Key Management
Usage Analytics & Budgets
Contact & Support
Resources
Troubleshooting
Frequently Asked Questions
Next Steps

Ecosystem

LLM Integrations

Cloud Platforms

Guardrails

Plugins

Vector Databases

Agents

AI Apps

Libraries

Tracing Providers

Strands Agents Documentation

Quick Start

How the Integration Works

Setting Up Portkey

Complete Integration Example

Production Features

1. Enhanced Observability

2. Reliability & Fallbacks

3. LLM Interoperability

Supported Providers

4. Guardrails for Safe Agents

Learn More About Guardrails

Advanced Configuration

Enterprise Governance

Centralized Key Management

Usage Analytics & Budgets

Contact & Support

Enterprise SLAs & Support

Portkey Community

Resources

Strands Agents Docs

Book a Demo

Troubleshooting

Frequently Asked Questions

Next Steps

Portkey Dashboard

Supported Models

Enterprise Features

Ecosystem

LLM Integrations

Cloud Platforms

Guardrails

Plugins

Vector Databases

Agents

AI Apps

Libraries

Tracing Providers

Strands Agents Documentation

​Quick Start

​How the Integration Works

​Setting Up Portkey

​Complete Integration Example

​Production Features

​1. Enhanced Observability

​2. Reliability & Fallbacks

​3. LLM Interoperability

Supported Providers

​4. Guardrails for Safe Agents

Learn More About Guardrails

​Advanced Configuration

​Enterprise Governance

​Centralized Key Management

​Usage Analytics & Budgets

​Contact & Support

Enterprise SLAs & Support

Portkey Community

​Resources

Strands Agents Docs

Book a Demo

​Troubleshooting

​Frequently Asked Questions

​Next Steps

Portkey Dashboard

Supported Models

Enterprise Features

Quick Start

How the Integration Works

Setting Up Portkey

Complete Integration Example

Production Features

1. Enhanced Observability

2. Reliability & Fallbacks

3. LLM Interoperability

4. Guardrails for Safe Agents

Advanced Configuration

Enterprise Governance

Centralized Key Management

Usage Analytics & Budgets

Contact & Support

Resources

Troubleshooting

Frequently Asked Questions

Next Steps