OpenRouter Model Catalog

Overview

This skill teaches you how to query the OpenRouter models API, filter by capabilities, and select the right model for your use case.

Prerequisites

• OpenRouter API key configured
• Basic understanding of LLM model differences

Instructions

Follow these steps to implement this skill:

1. Verify Prerequisites: Ensure all prerequisites listed above are met
2. Review the Implementation: Study the code examples and patterns below
3. Adapt to Your Environment: Modify configuration values for your setup
4. Test the Integration: Run the verification steps to confirm functionality
5. Monitor in Production: Set up appropriate logging and monitoring

Overview

This skill teaches you how to query the OpenRouter models API, filter by capabilities, and select the right model for your use case.

Prerequisites

• OpenRouter API key configured
• Basic understanding of LLM model differences

Listing All Models

API Request

curl https://openrouter.ai/api/v1/models \
  -H "Authorization: Bearer $OPENROUTER_API_KEY"

Python

import requests

response = requests.get(
    "https://openrouter.ai/api/v1/models",
    headers={"Authorization": f"Bearer {api_key}"}
)

models = response.json()["data"]
for model in models:
    print(f"{model['id']}: ${model['pricing']['prompt']}/1K tokens")

Model Categories

OpenAI Models

openai/gpt-4o              - Latest multimodal
openai/gpt-4-turbo         - Fast GPT-4
openai/gpt-4               - Original GPT-4
openai/gpt-4-32k           - Extended context
openai/gpt-3.5-turbo       - Fast and cheap
openai/gpt-3.5-turbo-16k   - Extended context

Anthropic Models

anthropic/claude-3.5-sonnet  - Best balance
anthropic/claude-3-opus      - Highest capability
anthropic/claude-3-sonnet    - Good balance
anthropic/claude-3-haiku     - Fastest Claude
anthropic/claude-2.1         - Legacy

Meta Llama Models

meta-llama/llama-3.1-405b-instruct  - Largest
meta-llama/llama-3.1-70b-instruct   - Great balance
meta-llama/llama-3.1-8b-instruct    - Fast/cheap
meta-llama/llama-3-70b-instruct     - Previous gen

Mistral Models

mistralai/mistral-large       - Most capable
mistralai/mistral-medium      - Balanced
mistralai/mixtral-8x7b-instruct - MoE model
mistralai/mistral-7b-instruct - Smallest

Google Models

google/gemini-pro-1.5      - Latest Gemini
google/gemini-pro          - Standard
google/palm-2-chat-bison   - PaLM 2

Specialized Models

# Code-focused
deepseek/deepseek-coder
codellama/codellama-34b-instruct

# Long context
anthropic/claude-3-opus (200K)
openai/gpt-4-turbo (128K)

# Cost-effective
meta-llama/llama-3.1-8b-instruct
mistralai/mistral-7b-instruct

Model Selection Criteria

By Task Type

Code generation:
  1. anthropic/claude-3.5-sonnet
  2. openai/gpt-4-turbo
  3. deepseek/deepseek-coder

Creative writing:
  1. anthropic/claude-3-opus
  2. openai/gpt-4
  3. meta-llama/llama-3.1-70b-instruct

Analysis/Reasoning:
  1. anthropic/claude-3-opus
  2. openai/gpt-4-turbo
  3. anthropic/claude-3.5-sonnet

Quick tasks:
  1. anthropic/claude-3-haiku
  2. openai/gpt-3.5-turbo
  3. mistralai/mistral-7b-instruct

By Context Length

< 4K tokens:   Any model
4K - 16K:      gpt-3.5-turbo-16k, most models
16K - 32K:     gpt-4-32k, claude-3 models
32K - 128K:    gpt-4-turbo, gemini-pro-1.5
128K - 200K:   claude-3-opus, claude-3.5-sonnet

By Cost (per 1M tokens)

Budget (< $1):
  - meta-llama/llama-3.1-8b-instruct
  - mistralai/mistral-7b-instruct

Mid-range ($1-5):
  - anthropic/claude-3-haiku
  - openai/gpt-3.5-turbo
  - meta-llama/llama-3.1-70b-instruct

Premium ($5-30):
  - anthropic/claude-3.5-sonnet
  - openai/gpt-4-turbo
  - anthropic/claude-3-sonnet

Enterprise ($30+):
  - anthropic/claude-3-opus
  - openai/gpt-4
  - meta-llama/llama-3.1-405b-instruct

Model Information API

Get Model Details

def get_model_info(model_id):
    response = requests.get(
        "https://openrouter.ai/api/v1/models",
        headers={"Authorization": f"Bearer {api_key}"}
    )
    models = response.json()["data"]
    return next((m for m in models if m["id"] == model_id), None)

model = get_model_info("anthropic/claude-3.5-sonnet")
print(f"Context: {model['context_length']}")
print(f"Prompt cost: ${model['pricing']['prompt']}/token")
print(f"Completion cost: ${model['pricing']['completion']}/token")

Model Schema

{
  "id": "anthropic/claude-3.5-sonnet",
  "name": "Claude 3.5 Sonnet",
  "description": "...",
  "context_length": 200000,
  "pricing": {
    "prompt": "0.000003",
    "completion": "0.000015"
  },
  "top_provider": {
    "context_length": 200000,
    "max_completion_tokens": 8192
  },
  "per_request_limits": null
}

Dynamic Model Selection

Based on Input Length

def select_model(input_tokens):
    if input_tokens < 4000:
        return "openai/gpt-3.5-turbo"
    elif input_tokens < 16000:
        return "openai/gpt-3.5-turbo-16k"
    elif input_tokens < 128000:
        return "openai/gpt-4-turbo"
    else:
        return "anthropic/claude-3.5-sonnet"

Based on Task

def select_model_for_task(task_type):
    models = {
        "code": "anthropic/claude-3.5-sonnet",
        "creative": "anthropic/claude-3-opus",
        "quick": "anthropic/claude-3-haiku",
        "analysis": "openai/gpt-4-turbo",
        "translation": "openai/gpt-4-turbo",
    }
    return models.get(task_type, "openai/gpt-4-turbo")

Model Comparison Tool

Compare Models Side-by-Side

async def compare_models(prompt, models):
    results = {}
    for model in models:
        response = await client.chat.completions.create(
            model=model,
            messages=[{"role": "user", "content": prompt}],
            max_tokens=500
        )
        results[model] = {
            "response": response.choices[0].message.content,
            "tokens": response.usage.total_tokens,
        }
    return results

# Usage
models = [
    "openai/gpt-4-turbo",
    "anthropic/claude-3.5-sonnet",
    "meta-llama/llama-3.1-70b-instruct"
]
results = await compare_models("Explain recursion", models)

Staying Updated

Model Changes

OpenRouter frequently adds new models:
- Check openrouter.ai/models for latest
- Subscribe to changelog
- API response includes all current models

Deprecation Handling

# Models may be deprecated
# Use fallbacks to handle gracefully
try:
    response = client.chat.completions.create(
        model="deprecated/model",
        messages=[...]
    )
except Exception as e:
    response = client.chat.completions.create(
        model="fallback/model",
        messages=[...]
    )

Output

Successful execution produces:

• Working OpenRouter integration
• Verified API connectivity
• Example responses demonstrating functionality

Error Handling

Common errors and solutions:

1. 401 Unauthorized: Check API key format (must start with sk-or-)
2. 429 Rate Limited: Implement exponential backoff
3. 500 Server Error: Retry with backoff, check OpenRouter status page
4. Model Not Found: Verify model ID includes provider prefix

Examples

See code examples in sections above for complete, runnable implementations.

Resources

• OpenRouter Documentation
• OpenRouter Models
• OpenRouter API Reference
• OpenRouter Status

Output

Successful execution produces:

• Working OpenRouter integration
• Verified API connectivity
• Example responses demonstrating functionality

Error Handling

Common errors and solutions:

1. 401 Unauthorized: Check API key format (must start with sk-or-)
2. 429 Rate Limited: Implement exponential backoff
3. 500 Server Error: Retry with backoff, check OpenRouter status page
4. Model Not Found: Verify model ID includes provider prefix

Examples

See code examples in sections above for complete, runnable implementations.

Resources

• OpenRouter Documentation
• OpenRouter Models
• OpenRouter API Reference
• OpenRouter Status