mxbai-embed-large-v1: local hardware requirements and GPU compatibility
← All models

Local AI model

mxbai-embed-large-v1

by Mixedbread AI

13 of 22 reference machines run it entirely in accelerator memory at 8k context, and 0 more with CPU offload.

335M Parameters 1k Maximum context 1 Quantization Full Compatibility coverage
Check my PC

Overview

Publisher
Mixedbread AI
Family
Mxbai
Series
mxbai Embed
Task
Embeddings
Architecture
cm.arch.encoder · bert
Parameters
335M
Maximum context
512 tokens
Licence
apache-2.0
Published
2024-03-07
Formats
gguf, safetensors
KV cache per token (f16)
0 B · no autoregressive KV cache (encoder)
Compatibility coverage
Full: verdicts computed
Upstream repository
mixedbread-ai/mxbai-embed-large-v1
Official page
https://www.mixedbread.com/

Quantizations and file sizes

Quantization Runtime Weights Memory floor at 4k Identity
F16gguf ollamaollama run mxbai-embed-large:latest 0.62 GB 0.62 GB digest verified

Sizes come from the runtime registry or repository listing. The memory floor adds the KV cache for 4,096 tokens when that figure cannot over-count; it is an estimate, not a requirement.

Runtimes

Compatibility on reference machines

8k context, 32 GB system RAM, best catalogued quantization per machine.

Machine Memory Answer Known facts
GeForce RTX 3060 Laptop GPU 6GB 6.00 GB VRAM Runs fully in memory estimated · F16 Weights 0.62 GB vs 6.00 GB VRAM: fitKV cache at 8k: no autoregressive KV cache (encoder)
GeForce RTX 4060 8GB 8.00 GB VRAM Runs fully in memory estimated · F16 Weights 0.62 GB vs 8.00 GB VRAM: fitKV cache at 8k: no autoregressive KV cache (encoder)
GeForce RTX 3080 10GB 10.0 GB VRAM Runs fully in memory estimated · F16 Weights 0.62 GB vs 10.0 GB VRAM: fitKV cache at 8k: no autoregressive KV cache (encoder)
GeForce RTX 3060 12GB 12.0 GB VRAM Runs fully in memory estimated · F16 Weights 0.62 GB vs 12.0 GB VRAM: fitKV cache at 8k: no autoregressive KV cache (encoder)
GeForce RTX 4070 12GB 12.0 GB VRAM Runs fully in memory estimated · F16 Weights 0.62 GB vs 12.0 GB VRAM: fitKV cache at 8k: no autoregressive KV cache (encoder)
GeForce RTX 5070 12GB 12.0 GB VRAM Runs fully in memory estimated · F16 Weights 0.62 GB vs 12.0 GB VRAM: fitKV cache at 8k: no autoregressive KV cache (encoder)
GeForce RTX 4060 Ti 16GB 16.0 GB VRAM Runs fully in memory estimated · F16 Weights 0.62 GB vs 16.0 GB VRAM: fitKV cache at 8k: no autoregressive KV cache (encoder)
GeForce RTX 4080 16GB 16.0 GB VRAM Runs fully in memory estimated · F16 Weights 0.62 GB vs 16.0 GB VRAM: fitKV cache at 8k: no autoregressive KV cache (encoder)
GeForce RTX 5060 Ti 16GB 16.0 GB VRAM Runs fully in memory estimated · F16 Weights 0.62 GB vs 16.0 GB VRAM: fitKV cache at 8k: no autoregressive KV cache (encoder)
GeForce RTX 5080 16GB 16.0 GB VRAM Runs fully in memory estimated · F16 Weights 0.62 GB vs 16.0 GB VRAM: fitKV cache at 8k: no autoregressive KV cache (encoder)
GeForce RTX 3090 24GB 24.0 GB VRAM Runs fully in memory estimated · F16 Weights 0.62 GB vs 24.0 GB VRAM: fitKV cache at 8k: no autoregressive KV cache (encoder)
GeForce RTX 4090 24GB 24.0 GB VRAM Runs fully in memory estimated · F16 Weights 0.62 GB vs 24.0 GB VRAM: fitKV cache at 8k: no autoregressive KV cache (encoder)
GeForce RTX 5090 32GB 32.0 GB VRAM Runs fully in memory estimated · F16 Weights 0.62 GB vs 32.0 GB VRAM: fitKV cache at 8k: no autoregressive KV cache (encoder)
Radeon RX 7600 8GB 8.00 GB VRAM Potential: weights fit not verified · F16 Weights 0.62 GB vs 8.00 GB VRAM: fitKV cache at 8k: no autoregressive KV cache (encoder)
Radeon RX 7800 XT 16GB 16.0 GB VRAM Potential: weights fit not verified · F16 Weights 0.62 GB vs 16.0 GB VRAM: fitKV cache at 8k: no autoregressive KV cache (encoder)
Radeon RX 9070 XT 16GB 16.0 GB VRAM Potential: weights fit not verified · F16 Weights 0.62 GB vs 16.0 GB VRAM: fitKV cache at 8k: no autoregressive KV cache (encoder)
Radeon RX 7900 XTX 24GB 24.0 GB VRAM Potential: weights fit not verified · F16 Weights 0.62 GB vs 24.0 GB VRAM: fitKV cache at 8k: no autoregressive KV cache (encoder)
Intel Arc B580 12GB 12.0 GB VRAM Not enough data not verified · F16 Weights 0.62 GB vs 12.0 GB VRAM: fitKV cache at 8k: no autoregressive KV cache (encoder)runtime support on this machine: unknown
Intel Arc A770 16GB 16.0 GB VRAM Not enough data not verified · F16 Weights 0.62 GB vs 16.0 GB VRAM: fitKV cache at 8k: no autoregressive KV cache (encoder)runtime support on this machine: unknown
Mac mini M4 24GB 24.0 GB unified Not enough data not verified · F16 Weights 0.62 GB vs 24.0 GB unified: fitKV cache at 8k: no autoregressive KV cache (encoder)
Mac mini M4 Pro 24GB 24.0 GB unified Not enough data not verified · F16 Weights 0.62 GB vs 24.0 GB unified: fitKV cache at 8k: no autoregressive KV cache (encoder)
MacBook Pro 14-inch M4 Max 36GB 36.0 GB unified Not enough data not verified · F16 Weights 0.62 GB vs 36.0 GB unified: fitKV cache at 8k: no autoregressive KV cache (encoder)

Definitive answers ("runs", "does not fit") come from the calibrated engine or a physical run. "Potential" rows only compare the weights with memory: the rest of the requirement is not modelled, so they are not verdicts. Unknown is never a failure.

Hardware for this model

Reference GPUs with 8 GB of VRAM or more run this model entirely in memory at 8k context: GeForce RTX 3060 Laptop GPU 6GB, GeForce RTX 4060 8GB, GeForce RTX 3080 10GB, GeForce RTX 3060 12GB.

  • Smallest catalogued artifact: 0.62 GB of weights (F16).
  • A GPU with at least 8 GB of VRAM holds those weights.
Explore 8 GB VRAM GPUs
These links open a hardware category search, not a specific product recommendation. Affiliate link — I may earn a commission from qualifying purchases. These links never change a compatibility answer. They follow the memory the data defends, not commission.

Sources and evidence

0 measured, 18 sourced, 5 estimated and 3 unknown fields.

Catalogue record verified 2026-09-16.

What else can your PC run?

Pick your GPU and system RAM to see every catalogued model that fits, with tested, estimated and potential answers kept apart.

Check my PC