Understand and compare
Llama 3 8B Instruct
vs.
Gemini Ultra
Overview
Llama 3 8B Instruct
|
Gemini Ultra
|
|
|---|---|---|
|
Provider
The entity that provides this model.
|
Meta
|
Google
|
|
Input Context Window
The number of tokens supported by the input context window.
|
8,000
tokens
|
32.8K
characters
|
|
Maximum Output Tokens
The number of tokens that can be generated by the model in a single request.
|
2,048
tokens
|
8,192
characters
|
|
Release Date
When the model was first released.
|
2024-04-18
|
Unknown
|
Pricing
|
Llama 3 8B Instruct
|
Gemini Ultra
|
|
|---|---|---|
|
Input
Cost of input data provided to the model.
|
Pricing not available.
|
Pricing not available.
|
|
Output
Cost of output tokens generated by the model.
|
Pricing not available.
|
Pricing not available.
|
Benchmarks
Compare relevant benchmarks between Llama 3 8B Instruct
and Gemini Ultra.
|
Llama 3 8B Instruct
|
Gemini Ultra
|
|
|---|---|---|
|
MMLU
Evaluating LLM knowledge acquisition in zero-shot and few-shot settings.
|
68.4
(5-shot)
|
83.7
(5-shot)
|
|
MMMU
A wide ranging multi-discipline and multimodal benchmark.
|
Benchmark not available.
|
59.4
(0-shot pass@1)
|
|
HellaSwag
A challenging sentence completion benchmark.
|
Benchmark not available.
|
Benchmark not available.
|
Llama 3 8B Instruct, developed by Meta, features a context window of 8000 tokens. The model was released on April 18, 2024, and achieved a score of 68.4 in the MMLU benchmark.
Gemini Ultra, developed by Google, features a large context window of 32768 tokens. The model has excelled in benchmarks like MMMU with a score of 59.4 in a 0-shot pass@1 scenario and MMLU with a score of 83.7 in a 5-shot scenario.
GPT-4 32K 0613
Mistral 8x7B Instruct
Claude 3 Haiku
Measure & Improve LLM
Product Performance.
Get Started
© 2023-2024 Context