Models/Llama 3.1 8B Instruct
MetaMeta / Llama 3.1 8B Instruct
Released: 7/23/2024
texttext
Input: $0.05 / Output: $0.08

meta-llama/Llama-3.1-8B-Instruct is an LLM designed for multilingual dialogue and general instruction-following tasks. It excels in handling conversational inputs, generating coherent text, and supporting multiple languages, making it suitable for chatbots, virtual assistants, and multilingual applications.

Some other noteworthy features of meta-llama/Llama-3.1-8B-Instruct include text generation and translation capabilities across at least eight major languages (English, German, French, Italian, Portuguese, Hindi, Spanish, and Thai) and efficient performance for both commercial and research use.

MetricValue
Parameter Count8 billion
Mixture of ExpertsNo
Context Length128,000 tokens
MultilingualYes
Quantized*No

*Quantization is specific to the inference provider and the model may be offered with different quantization levels by other providers.

Meta models available on Oxen.ai
ModalityPrice (1M tokens)
ModelInputOutputInputOutput
texttext$3.00$3.00
texttext$0.90$0.90
texttext$0.05$0.08
texttextN/AN/A
texttext$0.02$0.02
texttext$0.90$0.90
text, imagetext$0.22$0.88
texttext$0.08$0.30
See all models available on Oxen.ai