Skip to main content
Meta AIReleased December 6, 2024

Llama 3.3 70B

Open Source#13 Arena Rank70B parameters

Llama 3.3 70B holds a solid spot in the Arena rankings at #13. Context window: 0.128K tokens.

Context

128K

Input

Free (open)

Key Specifications

🏆

Arena Rank

#13

📐

Context Window

128K

📥

Input Price

per 1M tokens

Free (open)

📤

Output Price

per 1M tokens

Free (open)

🧠

Parameters

70B

🔓

Open Source

Yes

Best For

Instruction followingcodingreasoning

About Llama 3.3 70B

Llama 3.3 70B, developed by Meta AI, is an efficiency-optimized open-source model with 70 billion parameters and a 128K token context window. The model delivers capability comparable to the much larger Llama 3.1 405B, achieving near-frontier performance at a fraction of the compute requirements. This efficiency breakthrough means organizations can deploy competitive AI capabilities on significantly less hardware. Llama 3.3 excels at instruction following, coding, and structured reasoning tasks. Free and open-source, it runs on standard enterprise GPU setups and has become the de facto choice for organizations needing powerful, self-hosted AI. Its strong multilingual support covers dozens of languages. Llama 3.3 70B ranks #13 on the Chatbot Arena leaderboard, demonstrating that careful training optimization can close the gap between mid-size and frontier-scale models.

Pricing per 1M tokens

Input Tokens

Free (open)

Output Tokens

Free (open)

Compare Llama 3.3 70B

See how Llama 3.3 70B stacks up against other leading AI models

Llama 3.3 70B vs Llama 3.1 405BCompare →Llama 3.3 70B vs Llama 3.2 90B VisionCompare →Llama 3.3 70B vs Llama 3.1 70BCompare →Llama 3.3 70B vs Llama 3.1 8BCompare →Llama 3.3 70B vs Llama 3 8BCompare →

Other Meta AI Models

Llama 3.1 405B
128K#9
Llama 3.2 90B Vision
128K#11
Llama 3.1 70B
128K#14
Llama 3.1 8B
128K#22
Llama 3 8B
8K

Other Top Models

Claude Opus 4Anthropic
#1
GPT-4oOpenAI
#2
GPT-o3OpenAI
#2
GPT-o1OpenAI
#3
DeepSeek R1DeepSeek
#3
Claude Sonnet 4Anthropic
#3

Explore More

Model Benchmarks →All Models →

Frequently Asked Questions

What is Llama 3.3 70B?▾
Llama 3.3 70B, developed by Meta AI, is an efficiency-optimized open-source model with 70 billion parameters and a 128K token context window. The model delivers capability comparable to the much larger Llama 3.1 405B, achieving near-frontier performance at a fraction of the compute requirements. This efficiency breakthrough means organizations can deploy competitive AI capabilities on significantly less hardware. Llama 3.3 excels at instruction following, coding, and structured reasoning tasks. Free and open-source, it runs on standard enterprise GPU setups and has become the de facto choice for organizations needing powerful, self-hosted AI. Its strong multilingual support covers dozens of languages. Llama 3.3 70B ranks #13 on the Chatbot Arena leaderboard, demonstrating that careful training optimization can close the gap between mid-size and frontier-scale models.
How much does Llama 3.3 70B cost?▾
Input pricing for Llama 3.3 70B is Free (open) per million tokens; output runs Free (open). Token-based pricing means you can scale up or down without a fixed commitment.
What is Llama 3.3 70B's context window?▾
The context window for Llama 3.3 70B is 128K tokens. That's the maximum amount of text you can feed into a single prompt, including system instructions, conversation history, and the actual query.
Is Llama 3.3 70B open source?▾
Llama 3.3 70B is fully open source. You can grab the weights, run it on your own hardware, and fine-tune it for specific tasks. That flexibility is a big deal for teams with strict data requirements.
What is Llama 3.3 70B best for?▾
The sweet spot for Llama 3.3 70B is: Instruction following, coding, reasoning. If your workload fits one of these categories, it's worth benchmarking against alternatives.