3.3 70B Instruct
Llama 3Balanced
Tool UseStructured Output
About this model
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...
Performance Tier
Balanced
3.3 70B Instruct is a balanced model from Llama 3 : strong performance at a reasonable price.
Strong cost-performance ratio. Reliable for most professional use cases without premium pricing.
Pricing
This model is included in Elosia plansEco
Minimal cost. Ideal for very high volume or simple tasks.
| Type | per 1M tokens |
|---|---|
| Input (prompt) | $0.100 |
| Output (completion) | $0.320 |
Capabilities
Context Length131K
Max Output Tokens16K
TokenizerLlama3
Inputtext
Outputtext
Release DateDecember 6, 2024
Benchmarks
General Intelligence
MMLU
86%
MMLU-Pro
68.9%
Mathematics
MATH-500
77%
Programming
HumanEval
88.4%
Reasoning
IFEval
86.4%
Recommended Use Cases
CodingGeneral ChatAnalysisTranslation
Strengths
- Strong 70B model matching Llama 3.1 405B performance
- Excellent instruction following (IFEval 86.4%)
- Open-weight — widely deployed and well-supported
- Good cost-to-performance ratio via API providers
Limitations
- Superseded by Llama 4 family for most use cases
- No multimodal capabilities
Resources
This model may use your data for training
