Meta: Llama 3.3 70B Instruct

MetaID: meta-llama/llama-3.3-70b-instruct

The Meta Llama 3.3 multilingual large language model (LLM) is a pretrained and instruction tuned generative model in 70B (text in/text out). The Llama 3.3 instruction tuned text only model is optimized for multilingual dialogue use cases and outperforms many of the available open source and closed chat models on common industry benchmarks. Supported languages: English, German, French, Italian, Portuguese, Hindi, Spanish, and Thai. [Model Card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_3/MODEL_CARD.md)

Pricing per 1M Tokens

Input (Prompt)$0.10
Output (Completion)$0.32
Cache ReadFree
Cache WriteFree
ImageN/A

Specifications

Context Length131K
Max Output Tokens16K
Input ModalitiesText
Output ModalitiesText
TokenizerLlama3
Instruct Typellama3
Top Provider Context131K
Top Provider Max Output16K
ModeratedNo

Compare this model

See how Meta: Llama 3.3 70B Instruct stacks up against other models.

vs Meta: Llama Guard 4 12Bvs Meta: Llama 4 Maverickvs Meta: Llama 4 Scoutvs Llama Guard 3 8Bvs Meta: Llama 3.3 70B Instruct (free)

More from Meta

Meta: Llama Guard 4 12B
Input$0.18
Context164K
Meta: Llama 4 Maverick
Input$0.15
Context1.0M
Meta: Llama 4 Scout
Input$0.08
Context328K
Llama Guard 3 8B
Input$0.02
Context131K
Meta: Llama 3.3 70B Instruct (free)
InputFree
Context66K

Last updated: March 23, 2026

First tracked: March 23, 2026