Ithy Logo

Comprehensive Guide to Google LLM API Pricing

Understand pricing structures and free tier options for Google's powerful language models.

google cloud ai services

Key Takeaways

  • Flexible Pricing Models: Google offers a range of pricing options, including pay-as-you-go and subscription-based tiers, catering to diverse usage needs.
  • Generous Free Tier: Developers and businesses can access substantial free usage limits, allowing for experimentation and development without initial costs.
  • Comprehensive Integration: Google’s LLM APIs seamlessly integrate with other Google services like Google Workspace and Vertex AI, enhancing functionality and ease of use.

Overview of Google LLM APIs

Google provides a suite of Large Language Models (LLMs) accessible through various platforms, each tailored to different use cases and priced accordingly. The primary platforms offering these models include Vertex AI, Gemini, Google One AI Premium, and integrations within Google Workspace applications such as Gmail, Docs, and Sheets.

Pricing Tiers and Models

1. Pay-As-You-Go Model

Google's LLM APIs predominantly operate on a pay-as-you-go basis, where costs are determined by usage metrics, particularly the number of tokens processed. This model is flexible and scalable, making it suitable for both small-scale applications and large enterprises.

Vertex AI

Vertex AI hosts various LLMs, allowing users to access models from Google and its partners. Pricing for Vertex AI follows a usage-based approach, with costs accruing based on the number of input and output tokens processed during API calls. Additional features like advanced analytics and extended support may incur extra costs.

Gemini Models

The Gemini family, including models like Gemini 1.5 Pro, offers advanced language capabilities. Pricing specifics for Gemini models are outlined below:

Gemini 1.5 Pro Pricing Details:
  1. Per Call Cost: $0.0112 per request.
  2. Token Costs:
    • Input Tokens (Prompt): $3.50 per 1 million tokens.
    • Output Tokens (Generated Responses): $10.50 per 1 million tokens.
  3. Lower-Tier Token Costs (Basic API Access):
    • Input Tokens: $0.50 per 1 million tokens.
    • Output Tokens: $1.50 per 1 million tokens.

2. Subscription-Based Models

For users seeking more consistent access and additional features, Google offers subscription-based services:

Google One AI Premium

Google One AI Premium provides access to more advanced AI features powered by LLMs. Priced at $19.99 per month, this subscription includes access to Google’s most capable AI models, enhanced usage limits, priority support, and integration capabilities with other Google services. This tier is ideal for businesses and power users requiring sustained and high-volume access to LLM functionalities.

3. Integration with Google Workspace

Some Google Workspace applications integrate LLM-powered AI features. These features are typically included within standard Workspace subscriptions, though the extent of access might vary depending on the specific edition of Workspace subscribed to. For instance, AI features in Google Docs may include advanced grammar checking, content generation, and summarization tools, enhancing productivity without additional costs beyond the Workspace subscription.

Free Tier Options

Google offers a generous free tier to enable developers and businesses to experiment with and develop applications using their LLM APIs without incurring initial costs. The free tier details are outlined below:

Free Tier for Gemini API

  • Rate Limits: Up to 60 requests per minute at no cost.
  • Token Limits:
    • Input Tokens: Free up to 1 million tokens per minute.
    • Output Tokens: Free up to 1 million tokens per minute.
  • Context Caching: Free storage up to 1 million tokens per hour.
  • Tuning Service: Available free of charge.
  • Access via AI Studio: Completely free in all available regions, subject to phone number verification and data usage agreements (with exceptions in UK/CH/EEA/EU).
  • Data Usage: Data submitted during free tier usage may be used for training purposes, enhancing model performance, except in regions like the UK, Switzerland, EEA, and EU where stricter data privacy laws apply.

Free Tier for Vertex AI

Vertex AI also provides a free tier, allowing users to explore and develop with various LLMs without immediate financial commitment. The specifics of the free tier include limited API calls and token usage, making it suitable for development and experimentation. Users can access a subset of Vertex AI features at no cost, enabling them to prototype and test their applications before scaling up.

Free Tier for Google Workspace AI Features

Within Google Workspace, certain AI features are available without additional cost beyond the Workspace subscription. These include automated suggestions in Google Docs, smart reply in Gmail, and data analysis tools in Sheets. The free tier supports basic usage, allowing users to leverage AI capabilities to enhance productivity and collaboration.

Detailed Pricing Comparison

Service Free Tier Pricing Details
Vertex AI
  • Limited free API calls
  • Limited token usage
  • Access to basic features
  • Input Tokens: $3.50 per 1M tokens
  • Output Tokens: $10.50 per 1M tokens
  • Per Call Cost: Varies based on usage
Gemini 1.5 Pro
  • Up to 60 requests per minute
  • 1M tokens per minute
  • Context caching up to 1M tokens per hour
  • Tuning service included
  • Per Call Cost: $0.0112 per request
  • Input Tokens: $3.50 per 1M tokens
  • Output Tokens: $10.50 per 1M tokens
  • Lower-Tier Access:
    • Input Tokens: $0.50 per 1M tokens
    • Output Tokens: $1.50 per 1M tokens
Google One AI Premium Not applicable
  • Subscription: $19.99 per month
  • Includes access to advanced models and features
  • Enhanced usage limits
  • Priority support
Google Workspace AI Features
  • Included with Workspace subscription
  • Basic AI features in Docs, Gmail, Sheets
  • Pricing varies based on Workspace edition
  • Advanced AI features may require higher-tier subscriptions
Gemini 1.5 Flash 15 requests per minute
  • Input Tokens: $0.075 per 1M tokens (up to 128k tokens)
  • Input Tokens: $0.15 per 1M tokens (for longer prompts)
  • Output Tokens: $0.30 per 1M tokens (up to 128k tokens)
  • Output Tokens: $0.60 per 1M tokens (for longer prompts)
Gemini 1.5 Flash-8B 1.5 requests per minute
  • Input Tokens: $0.0375 per 1M tokens (up to 128k tokens)
  • Input Tokens: $0.075 per 1M tokens (for longer prompts)
  • Output Tokens: $0.15 per 1M tokens (up to 128k tokens)
  • Output Tokens: $0.30 per 1M tokens (for longer prompts)

Additional Considerations

Optimizing Costs

To manage and optimize costs when using Google's LLM APIs, consider selecting lower-tier access or opting for smaller-scale versions of the models based on specific needs. Efficient token management, such as minimizing unnecessary token usage and leveraging context caching, can significantly reduce expenses. Additionally, monitoring usage through Google Cloud’s billing tools and setting up alerts can help prevent unexpected costs.

Integration with Other Google Services

Google's LLM APIs are designed to integrate seamlessly with other Google services such as Gmail, Google Docs, Sheets, and Vertex AI. This integration enhances productivity and allows for more advanced applications, like automated content generation, summarization, and language translation within familiar Google environments. Leveraging these integrations can streamline workflows and add value to existing tools without requiring extensive development efforts.

Data Privacy and Usage

While using the free tier, it is important to note that data submitted to Google's APIs may be used for training purposes, enhancing model performance, except in regions like the UK, Switzerland, EEA, and EU where stricter data privacy laws apply. Developers should review and comply with data usage agreements, especially when handling sensitive or confidential information. Opting out of data usage for training may be necessary for certain applications to ensure compliance with regional regulations.

Conclusion

Google offers a robust and flexible pricing structure for its LLM APIs, accommodating a wide range of users from individual developers to large enterprises. With generous free tiers and various pay-as-you-go and subscription options, Google ensures that users can effectively leverage advanced language models while managing costs and scalability. Additionally, seamless integration with other Google services enhances the utility and ease of adoption, making Google's LLM APIs a compelling choice for diverse language processing needs.

References


Last updated January 28, 2025
Search Again