Llama 3 Instruct 70B: Intelligence, Performance & Price Analysis
Analysis of Meta's Llama 3 Instruct 70B and comparison to other AI models across key metrics including quality, price, performance (tokens per second & time to first token), context window & more. Click on any model to compare API providers for that model. For more details including relating to our methodology, see our FAQs.
Note: Some providers are deprecating their Llama 3 endpoints in favour of Llama 3.1 endpoints
Meta has launched a newer model, Llama 3.1 70B. We suggest considering this model instead of Llama 3 70B. See the following pages for a comparison of Llama 3.1 70B to other models and Llama 3.1 70B API provider benchmarks.
Comparison Summary
Intelligence:
Llama 3 70B is of lower quality compared to average, with a MMLU score of 0.574 and a Intelligence Index across evaluations of 27.
Price:Llama 3 70B is cheaper compared to average with a price of $0.84 per 1M Tokens (blended 3:1).
Llama 3 70B Input token price: $0.73, Output token price: $0.84 per 1M Tokens.
Speed:Llama 3 70B Input token price: $0.73, Output token price: $0.84 per 1M Tokens.
Llama 3 70B is slower compared to average, with a output speed of 34.6 tokens per second.
Latency:Llama 3 70B has a lower latency compared to average, taking 0.69s to receive the first token (TTFT).
Context Window:Llama 3 70B has a smaller context windows than average, with a context window of 8.2k tokens.
Highlights
Intelligence
Artificial Analysis Intelligence Index; Higher is better
Speed
Output Tokens per Second; Higher is better
Price
USD per 1M Tokens; Lower is better
Parallel Queries:
Prompt Length:
Note: Long prompts not supported as a context window of at least 10k tokens is required
Llama 3 Instruct 70B Model Details
Comparisons to Llama 3 70B
Claude 4 Sonnet
Claude 4 Sonnet (Extended Thinking)
o4-mini (high)
GPT-4.1
o3
Llama 4 Maverick
Llama 4 Scout
Gemini 2.5 Flash Preview (May '25) (Reasoning)
Gemini 2.5 Pro Preview (Mar' 25)
Claude 3.7 Sonnet (Extended Thinking)
Mistral Medium 3
DeepSeek R1
DeepSeek V3 0324 (Mar' 25)
Grok 3 mini Reasoning (high)
Nova Premier
Qwen3 235B A22B (Reasoning)
GPT-4o (Nov '24)
Further details