Find the AI option that gets the job done for less

Compare any two of three editable routes: hosted API, local or self-hosted, or rented GPU. See estimated monthly cost and the break-even point, then include retries and recovery when reliability matters.

Open the LLM cost calculator

No account required · private in your browser · exportable results

How it works

1. Describe the task and monthly volume.
2. Model two execution routes.
3. Inspect monthly cost, break-even, retries, and recovery.

Find a model → · Best by use → · Compare models → · View price trends →

Popular: models for 24 GB VRAM · best LLMs for 12 GB VRAM · laptop LLMs · speech-to-text models.

Popular models

• Cost routes for Llama 3.1 8B Instruct
• Cost routes for Llama 3.1 70B Instruct
• Cost routes for Llama 3.1 405B Instruct
• Cost routes for Qwen2.5 7B Instruct
• Cost routes for Qwen2.5-Coder 32B Instruct
• Cost routes for Mistral 7B Instruct v0.3
• Cost routes for Mixtral 8x7B Instruct
• Cost routes for Gemma 2 9B Instruct