The AI model Gemini 1.5 Flash has been released in a new version. Gemini Flash-8B is more compact, faster and cheaper than its predecessor.
Gemini 1.5 Flash-8B is available as of now and is said to be one of the most affordable lightweight LLMs on the market. Optimized for speed and efficiency, the model was developed primarily to run on devices such as smartphones and sensors. The AI workload is smaller because that hardware does not allow for high performance.
Good performance for compact model
Still, this lighter version should not be inferior to its predecessors. According to benchmarks tests, it even delivers comparable performance in some areas. Tasks like chatting, transcription and contextually correct translation of long texts are no problem, SiliconANGLE knows.
Gemini 1.5 Flash was announced at Google I/O in May 2024, and was made available to paying customers a few weeks later. At the time of release, Google reported that the input size was 60 times that of OpenAI’s GPT-3.5 Turbo. Flash-8B is another double the speed. The number of requests per minute also doubled to 4,000 compared to 1.5 Flash.
In terms of price, Gemini 1.5 Flash-8B is around the recommended price of similar models from OpenAI and Anthropic PBC. Flash-8B costs $0.15 per one million exhaust tokens and only $0.01 per one million reused input tokens. At OpenAI, the cheapest model is GPT-4o mini at $0.15 per one million input tokens. That gets half as cheap if you reuse prompt prefixes or work with batches. In contrast, Claude 3 Haiku of Anthropic charges $0.25 per one million input tokens, and $0.03 per one million input tokens for reused tokens.
read also