MetaLlama 3.3Open SourceArena #154Dec 6, 2024
Meta: Llama 3.3 70B Instruct (free)
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
Arena Rank
#154
of 305 models
Output Speed
175
tokens/sec
Time to First Token
320ms
TTFT
Supported Parameters
Frequency Penalty
Reduce repetition
Max Tokens
Output length limit
Presence Penalty
Encourage new topics
Seed
Deterministic outputs
Stop Sequences
Custom stop tokens
Temperature
Controls randomness
Tool Choice
Control tool usage
Tool Calling
Function calling support
Pricing Comparison
| Router | Input / 1M | Output / 1M | Cached Input / 1M |
|---|
| OpenRouter | Free | Free | — |
| Vercel AI | $0.72 | $0.72 | — |
| Martian | $0.10 | $0.32 | — |
Benchmarks
Open LLM Leaderboard
AverageOpen LLM Leaderboard
44.85/100IFEvalOpen LLM Leaderboard
89.98/100BBHOpen LLM Leaderboard
56.56/100MATH Lvl 5Open LLM Leaderboard
48.34/100GPQAOpen LLM Leaderboard
10.51/100MUSROpen LLM Leaderboard
15.57/100MMLU-PROOpen LLM Leaderboard
48.13/100Artificial Analysis
Intelligence IndexArtificial Analysis
48/100Coding IndexArtificial Analysis
42/100Math IndexArtificial Analysis
50/100MMLU-PROArtificial Analysis
0.658/1GPQA DiamondArtificial Analysis
0.448/1MATH-500Artificial Analysis
0.712/1AIME 2024Artificial Analysis
0.067/1LiveCodeBenchArtificial Analysis
0.365/1Aider Code Editing
Aider EditingAider Code Editing
42.1/100Model IDs
OpenRoutermeta-llama/llama-3.3-70b-instruct:free
Similar Models
Ranked by provider, pricing, capabilities, and arena performance