A rate per million tokens is not a budget. The calculator takes your request volume, prompt size and cache hit rate and gives you a monthly figure.
Llama 3.3 70B (Groq) API pricing (October 2026)
Retired from Groq self-serve on 16 Aug 2026; enterprise contracts only. Last published price.
What it costs at real volumes
A chat-shaped workload: 800 input tokens and 300 output tokens per request, no caching, no batch.
Specifications
- Provider
- Groq
- Context window
- 128K tokens
- Max output
- 33K tokens
- Tier
- Workhorse
- Status
- Retired
- Blended $/1M (3:1)
- $0.640
- Cache break-even, caching each turn
- No caching
- Prices verified
- 12 Oct 2026
Source: Groq pricing. Finitizer is not affiliated with Groq.
Similar models
Other Llama models
- Llama 3.1 8B (Groq)-$0.583
Closest prices elsewhere
- GPT-5.6 Luna-$0.190
- Gemini 3.5 Flash-Lite+$0.210
- Amazon Nova 2 Lite+$0.210
- DeepSeek V4 Pro+$0.350
Frequently asked questions
How much did Llama 3.3 70B (Groq) cost per 1M tokens?
Llama 3.3 70B (Groq) cost $0.590 per million input tokens and $0.790 per million output tokens at Groq final list prices. Output was 1.3x the input rate, so the input:output ratio of a workload drives the real cost more than the headline price.
Does Llama 3.3 70B (Groq) support prompt caching?
Llama 3.3 70B (Groq) has no published prompt-caching rate, so repeated context is billed at the full input rate on every request.
What is the batch discount for Llama 3.3 70B (Groq)?
Requests submitted through the asynchronous batch endpoint are billed at 50% off both input and output rates, bringing input to $0.295 and output to $0.395 per million tokens.
What is Llama 3.3 70B (Groq)'s context window?
128K tokens of context with up to 33K output tokens per request. A larger context window raises the ceiling on what you can send, not the price per token, but filling it on every request is one of the most common causes of an unexpected bill.
Running Llama 3.3 70B (Groq) in production?
Finitizer Tokenomics reads what this model actually costs you from your own bills and exports, shows which workload, team or API key is spending it, flags a spike or a new spend source the day your data shows it, and forecasts where the month will land.
See Finitizer Tokenomics