The fast, cheap tier of Gemini 3.6. Flash keeps a 1M token window and multimodal input while running several times faster than Pro, which makes it the default choice for high-volume production traffic.
Use our main calculator for more detailed estimates including input/output combinations.
Key Features
1M token context window
Very high throughput
Multimodal input (text, image, audio, video)
Context caching
Thinking budget control
Common Use Cases
High-volume APIs
Real-time assistants
Media pipelines
RAG front ends
Ratings & Feedback
0.0 / 5 · 0 votes
Comments
Frequently Asked Questions
What are Google's main AI model families as of late 2024?
Google's flagship AI models are part of the Gemini family. This includes the highly capable Gemini 1.5 Pro, the faster Gemini 1.5 Flash, and potentially newer, more advanced versions like Gemini 2.0 Ultra (or a similar successor). These models are designed for strong multimodality and advanced reasoning, often featuring extremely large context windows.
What is a key differentiating feature of Google's Gemini 1.5 Pro and newer Gemini models?
A standout feature of Gemini 1.5 Pro and its successors is their extremely large context window, often starting at 1 million tokens and potentially extending to 2 million tokens or more. This allows for in-depth analysis of vast amounts of information, such as entire codebases, multiple long documents, or hours of video/audio content, in a single pass.
How does Google price its Gemini models, especially those with very large context windows?
Google Gemini models are typically priced per 1,000 tokens (or sometimes per 1,000 characters for certain APIs/regions). For models with very large context windows like Gemini 1.5 Pro, pricing might be tiered, with a lower rate for usage within a standard context (e.g., up to 128K tokens) and a higher rate for processing tokens beyond that threshold, up to the model's maximum capacity.
Are Google's Gemini models primarily focused on multimodality?
Yes, multimodality is a core strength of the Gemini family. They are designed from the ground up to understand, combine, and generate information across various data types including text, code, images, audio, and video. This makes them suitable for a wide range of complex tasks that require processing diverse inputs.