Skip to content

Phi 3.5 Mini Instruct on RTX 5090: VRAM, speed and fit

Phi 3.5 Mini Instruct at 8-bit runs on the RTX 5090 with 64K of context at about 243 tokens a second.

Text generation · Chat · Ollama / LM Studio

Runs comfortably · on RTX 5090 · 8-bit (Q8_0) · Ollama / LM Studio · Chat

GPU memory
28.4 GB
of 32.0 GB
Context
64K
auto
Speed
243 tok/s
41 when full
First reply
0.1 s
2K prompt

Where the memory goes

Model
3.8 GB
Context
24.0 GB
Overhead
619 MB
OS
0.0 GB
Free
3.6 GB
Memory
28.4 GB of 32.0 GB
Context
64K tokens · auto
Speed
243 tok/s · 41 when full
First reply
0.1 s for a 2K prompt
Running cost
$9.30/mo at home · $58/mo on a cloud 2× RTX 2000 Ada

Validation: Validated

What Phi 3.5 Mini Instruct on RTX 5090 was checked against
CheckTargetResultStatus
Fits in GPU memory32.0 GB usable28.4 GB neededPass
Headroom5% free3.6 GB free (11%)Pass
Context for chat16K or more64KPass
Writing speed15 tok/s or more243 tok/sPass
First replyunder 3 s, 2K prompt0.1 sPass

Targets: a reply you can read along with, a first reply the use case tolerates, and 5% of memory left free. Calculated from published specs, not benchmarked.

The setup

Model
Phi 3.5 Mini Instruct
Parameters
3.8B
Precision
8-bit (Q8_0)
Context
64K, auto
Chats at once
1
Use
Chat
Runs in
Ollama / LM Studio
Hardware
RTX 5090
GPU can use
32.0 GB of 32.0 GB

Running cost

At home
$9.30/mo
425 W at the wall
Renting instead
$58/mo
2× RTX 2000 Ada at $0.48/h

4 h a day, power at $0.18 per kWh, cloud at RunPod on-demand prices as of Nov 22, 2026.

How to run it

llama.cpp
llama-server -hf bartowski/Phi-3.5-mini-instruct-GGUF
vLLM
vllm serve microsoft/Phi-3.5-mini-instruct

Phi 3.5 Mini Instruct on similar machines

All machines: Phi 3.5 Mini Instruct VRAM requirements. Everything that runs here: what can the RTX 5090 run?

Other models that run on RTX 5090

Questions

Can Phi 3.5 Mini Instruct run on RTX 5090?

Phi 3.5 Mini Instruct at 8-bit runs on the RTX 5090 with 64K of context at about 243 tokens a second.

Report VR-02D9CIT, calculated Nov 22, 2026 from published specs, not benchmarked.