Google Gemini Models

Explore the Google Gemini language and embedding models available through our OpenAI Assistants API-compatible service.

Google Gemini logo

Google: Gemini Pro 1.5 Experimental

Context Length:
1,000,000 tokens
Architecture:
text+image->text
Max Output:
8,192 tokens

Pricing:

Gemini 1.5 Pro Experimental represents the cutting-edge iteration of the Gemini 1.5 Pro model. As an experimental release, this version is subject to strict rate-limiting by Google to ensure optimal performance and stability during testing.

Please note that access to and usage of Gemini 1.5 Pro Experimental is governed by Google's Gemini Terms of Use. This advanced model supports multimodal capabilities, enabling seamless integration across diverse data types and applications.

Google: Gemini Flash 1.5 Experimental

Context Length:
1,000,000 tokens
Architecture:
text+image->text
Max Output:
8,192 tokens

Pricing:

Gemini 1.5 Flash Experimental is a cutting-edge, experimental iteration of the Gemini 1.5 Flash model. Designed for exploration and testing, this version is not intended for production environments. Please note that its availability is subject to change, and it may be updated, replaced, or redirected to another model in the future.

Usage of Gemini is governed by Google's Gemini Terms of Use.

Important: This model is strictly experimental and should not be used for production-level applications.

Google: Gemini Flash 1.5 8B Experimental

Context Length:
1,000,000 tokens
Architecture:
text+image->text
Max Output:
8,192 tokens

Pricing:

Introducing Gemini Flash 1.5 8B Experimental, a cutting-edge, experimental model featuring 8 billion parameters, derived from the Gemini Flash 1.5 architecture. Designed for exploration and innovation, this multimodal model showcases advanced capabilities but is currently in a testing phase.

Please note that Gemini Flash 1.5 8B Experimental is not recommended for production environments and may be subject to significant rate limitations. Usage of this model is governed by Google's Gemini Terms of Use.

Explore the potential of this experimental model while keeping in mind its developmental status and limitations.

Google: Gemini 2.0 Flash Thinking Experimental (free)

Context Length:
40,000 tokens
Architecture:
text+image->text
Max Output:
8,000 tokens

Pricing:

Gemini 2.0 Flash Thinking Mode is an innovative experimental model designed to generate and display the internal "thinking process" it undergoes while formulating responses. This advanced feature enhances the model's reasoning capabilities, enabling it to deliver more insightful and logically structured answers compared to the standard Gemini 2.0 Flash model. By revealing its cognitive steps, Thinking Mode offers users a deeper understanding of how conclusions are reached, making it a powerful tool for complex problem-solving and decision-making.

Google: Gemini Flash 2.0 Experimental (free)

Context Length:
1,048,576 tokens
Architecture:
text+image->text
Max Output:
8,192 tokens

Pricing:

Gemini Flash 2.0 delivers a dramatically improved time to first token (TTFT) compared to its predecessor, Gemini Flash 1.5, while maintaining the high-quality performance of larger models such as Gemini Pro 1.5. This latest version introduces significant upgrades in multimodal understanding, coding proficiency, complex instruction execution, and function calling. Together, these enhancements create a more seamless, powerful, and reliable agentic experience for users.

Google: Gemini Flash 1.5 8B

Context Length:
1,000,000 tokens
Architecture:
text+image->text
Max Output:
8,192 tokens

Pricing:

Prompt: $0.0000000375
Completion: $0.00000015

The Gemini Flash 1.5 8B model is engineered for exceptional speed and efficiency, delivering superior performance in tasks such as chat, transcription, and translation, particularly for small prompts. With significantly reduced latency, it excels in real-time applications and large-scale operations, ensuring seamless responsiveness. Designed with cost-effectiveness in mind, this model maintains high-quality outputs while optimizing resource utilization.

Explore more about this model here.

Use of Gemini is governed by Google's Gemini Terms of Use.

Google: Gemini Flash 1.5

Context Length:
1,000,000 tokens
Architecture:
text+image->text
Max Output:
8,192 tokens

Pricing:

Prompt: $0.000000075
Completion: $0.0000003
Image: $0.00004

Gemini 1.5 Flash is a cutting-edge foundation model engineered to excel in a wide range of multimodal tasks, including visual understanding, classification, summarization, and content generation from images, audio, and video. It seamlessly processes both visual and textual inputs, such as photographs, documents, infographics, and screenshots, delivering versatile and accurate results.

Optimized for high-volume, high-frequency operations, Gemini 1.5 Flash prioritizes cost-efficiency and low latency without compromising performance. It achieves quality comparable to other Gemini Pro models on most common tasks, but at a significantly reduced cost. This makes it an ideal solution for applications like chat assistants and on-demand content generation, where speed, scalability, and affordability are critical.

Usage of Gemini 1.5 Flash is governed by Google's Gemini Terms of Use.

Google: Gemini Pro 1.0

Context Length:
32,760 tokens
Architecture:
text->text
Max Output:
8,192 tokens

Pricing:

Prompt: $0.0000005
Completion: $0.0000015
Image: $0.0025

Introducing Google's premier text generation model, Gemini. This advanced AI is expertly designed to manage a wide range of natural language tasks, including multiturn text and code chat, as well as code generation.

For detailed benchmarks and comprehensive prompting guidelines, visit Deepmind.

Please note that the use of Gemini is governed by Google's Gemini Terms of Use.

Google: Gemini Pro 1.5

Context Length:
2,000,000 tokens
Architecture:
text+image->text
Max Output:
8,192 tokens

Pricing:

Prompt: $0.00000125
Completion: $0.000005
Image: $0.0006575

Introducing Google's cutting-edge multimodal model, designed to seamlessly integrate image and video inputs within text or chat prompts. This advanced tool is optimized for a wide range of language-based tasks, including:

  • Code generation
  • Text generation and editing
  • Problem solving
  • Personalized recommendations
  • Information and data extraction
  • Data generation
  • AI agent development

Please note that the use of Gemini is governed by Google's Gemini Terms of Use.

*Currently, video input functionality is not supported via OpenRouter.

Ready to build with Google Gemini?

Start using these powerful models in your applications with our flexible pricing plans.