GPU VRAM calculator

Run d1-3B locally: hardware requirements

d1-3B needs about 10.8 GB of GPU memory for one person at 32K context, and about 3.8 GB more for each extra person at 64K. For 80 users working as AI agents, plan on 38.9 GB, which fits one L40S.

Liquid AI, United States. From its files on Hugging Face, checked 2026-11-05.

Context per person: more context lets the model work with longer conversations and more files, and needs more GPU memory for every person. As a guide: 8K for short chat, 32K for chat over documents, 64K for AI agents, whose instructions, tools and files all count.

80 users means about 16 people working at the same moment. That share is a planning assumption; if you know yours, choose the total that matches it.

Work it out for your team: LiquidAI/d1-3B

On Hugging Face

Liquid AI, United States · 3.1B parameters · License: other

What you need for 16 active users

1 × L40S

It holds the model and up to 19 active users at 64K context.

The server

128 GB RAM
48 CPU cores
250 GB NVMe

Rules of thumb

GPU memory in total
38.9 GB
Weights
6.2 GB (published files)
Conversation cache, all active
30.0 GB
Engine (vLLM)
3.0 GB

Your options by card

80 users means about 16 using the model at any moment. The others are not waiting; they are simply doing something else. If more become active than a build handles, they queue.

CardMemoryActive users, one serverFor your 16 active
L40S48 GB191 × L40S
RTX 6000 Ada48 GB191 × RTX 6000 Ada
RTX PRO 5000 Blackwell48 GB191 × RTX PRO 5000 Blackwell
A10080 GB351 × A100
H10080 GB351 × H100
RTX PRO 6000 Blackwell96 GB431 × RTX PRO 6000 Blackwell
H200 NVL141 GB661 × H200 NVL
B200180 GB861 × B200
GeForce RTX 309024 GB192 × GeForce RTX 3090
GeForce RTX 409024 GB192 × GeForce RTX 4090
GeForce RTX 507012 GB73 servers, each with 2 × GeForce RTX 5070
GeForce RTX 508016 GB112 servers, each with 2 × GeForce RTX 5080
GeForce RTX 509032 GB272 × GeForce RTX 5090
RTX 4000 Ada20 GB253 × RTX 4000 Ada
RTX A500024 GB192 × RTX A5000
RTX PRO 4000 Blackwell24 GB192 × RTX PRO 4000 Blackwell
RTX PRO 4500 Blackwell32 GB272 × RTX PRO 4500 Blackwell

These are memory limits. A card can hold more people than it can answer quickly, so test your own workload before you buy. More than one card in a server needs an engine that splits the model.

d1-3B by team size

AI agents, 1 user in 5 active at a time, 64K context, as published.

Total usersAt onceGPU memoryWhat you need
20416.4 GB1 × RTX 4000 Ada
801638.9 GB1 × L40S
1202453.9 GB1 × A100
500100197 GB2 × H200 NVL

All models

How this is worked out

GPU memory = weights + a conversation cache for each person + the serving engine, from this model's files on Hugging Face. These are memory limits, not speed. The full method and its limits.

Questions

How much GPU memory does d1-3B need?

d1-3B needs about 10.8 GB of GPU memory for one person at 32K context, and about 3.8 GB more for each extra person at 64K. For 80 users working as AI agents, plan on 38.9 GB, which fits one L40S.

How many active users can d1-3B handle on one 96 GB GPU?

By memory, about 43 active users at once at 64K context on a card such as the RTX PRO 6000 Blackwell. Speed can set a lower limit, so load-test before you buy.

Can d1-3B run on a 24 GB GPU such as the RTX 4090?

Yes as published for one person at 32K context: it needs about 10.8 GB.

What license does d1-3B use?

Hugging Face lists its license as other. Check the license on the model's page before you use it.

Running a model on your own servers?

NeueCode is private autonomous intelligence for enterprise: it takes on your organization's tasks on your own GPUs, with nothing leaving your network, a person approving every action, and all of it on record. It runs each model on one card, so read the single-card figures above.

What you need to run NeueCode