Groq API Pricing Calculator
Estimate your monthly and annual token costs for Groq’s high-speed LPU inference engine in 2026. Calculate costs dynamically based on model choice, input/output volumes, and Whisper audio transcription times.
Your Monthly Workload
Configure model selection and monthly API utilization
Pricing varies by compute complexity and context handling capabilities.
The amount of context, instructions, and history sent to the Groq API.
The response text generated by the model on Groq LPUs.
Cost Estimation
Annual & Monthly budget breakdown
Enter your expected text token volume or audio transcription duration, select your model, and click Calculate to inspect your custom budget matrix.
Groq API Model Specifications
Comprehensive pricing list of available LLMs and transcription models on Groq’s high-efficiency LPU infrastructure in 2026.
| Model Name | Input Rate (Per 1M) | Output Rate (Per 1M) | Billing Type |
|---|---|---|---|
| Llama 3.3 70B (Versatile) | $0.59 | $0.79 | Per-Token Pay-As-You-Go |
| Llama 3.1 70B (Legacy Large) | $0.59 | $0.79 | Per-Token Pay-As-You-Go |
| DeepSeek R1 Distill Llama 70B | $0.59 | $0.79 | Per-Token Pay-As-You-Go |
| Mixtral 8x7B (MoE) | $0.24 | $0.24 | Per-Token Pay-As-You-Go |
| Qwen 2.5 Coder 32B | $0.39 | $0.39 | Per-Token Pay-As-You-Go |
| Gemma 2 9B (Google) | $0.20 | $0.20 | Per-Token Pay-As-You-Go |
| Llama 3.1 8B (Instant) | $0.05 | $0.08 | Per-Token Pay-As-You-Go |
| Whisper Large V3 | $0.01 / min | N/A | Per Transcription Minute |
Groq API FAQ
Everything you need to know about understanding, calculating, and optimizing your high-speed LPU budget.
Groq is an ultra-fast AI inference platform powered by custom silicon called the Language Processing Unit (LPU). It is designed to serve leading open-source models like Llama 3, Mixtral, and Gemma at record-breaking real-time generation speeds.
Groq API operates on a transparent, pay-as-you-go model with no monthly subscription base fees. You are billed strictly per million tokens processed, separate for inputs (prompts) and outputs (completions). For audio models like Whisper, you are billed per minute of recorded transcription.
In 2026, the Llama 3.3 70B Versatile model on Groq is priced at $0.59 per million input (prompt) tokens and $0.79 per million output (completion) tokens.
Llama 3.1 8B is one of the most cost-effective models available, priced at just $0.05 per million input tokens and $0.08 per million output tokens on Groq’s LPU hardware.
Whisper Large V3 is priced on a transcription-minute basis at $0.01 per minute, making it exceptionally cost-efficient for heavy audio workloads.
As a rule of thumb, 1 token is equivalent to approximately 4 characters of English text or 0.75 words. Therefore, 1 million tokens equals roughly 750,000 words of conversational material.
No. There are no monthly base plans, minimum spending commitments, or setup fees for using the standard Groq API. You pay only for what you actually consume.
To reduce costs, implement prompt caching, limit the context size passed into the system, avoid redundant prompt templates, and consider utilizing smaller models like Llama 3.1 8B for simpler categorization or retrieval tasks.