One job, one card, one API
Mistral: Mistral Small 4 on GeForce RTX 3090
The sums below use the “Chatbot for a small team” example: 4,200 jobs a month, 1,500 tokens in, 400 out. 8 people, 25 questions on each of 21 working days. 1,500 tokens in (the question plus a short history) and 400 tokens out. A worked example, not a log from a real team.
Check the figures
Mistral: Mistral Small 4 will not fit on GeForce RTX 3090. Even the smallest quant on the ladder, Q2_K, is about 44.8 GB against the card's 24 GB. Choose a bigger card or a smaller model. This site refuses the pairing rather than quote a speed it cannot run.