Skip to main content
ModelLLMsvLlama-3.3-70B-Instruct

Llama 3.3 70B

by Meta · open-source · Last verified 2026-03-17

Meta's refined 70B model delivering performance comparable to the much larger 405B variant through improved training techniques. Offers the best performance-to-cost ratio in the Llama family.

https://llama.meta.com
B
BAbove Average
Adoption: AQuality: AFreshness: B+Citations: B+Engagement: F

Specifications

License
Llama 3.3 Community License
Pricing
open-source
Capabilities
text-generation, code-generation, reasoning, tool-use, long-context, multilingual
Integrations
huggingface, ollama, vllm, together-ai, aws-bedrock, groq
Use Cases
enterprise-deployment, code-generation, research, agentic-workflows, content-creation
API Available
No
Parameters
70B
Context Window
128K tokens
Modalities
text
Training Cutoff
December 2024
Tags
llm, open-source, large-model, cost-efficient, reasoning, meta
Added
2026-03-17
Completeness
100%

Index Score

68.95
Adoption
82
Quality
87
Freshness
78
Citations
75
Engagement
0

Explore the full AI ecosystem on Agents as a Service