Home / Machines / CPU only 32GB

What AI can a PC with no GPU and 32 GB RAM run?

PC without a graphics card, 32 GB RAM. Updated 2 November 2026.

Quick verdict

PC without a graphics card, 32 GB RAM can run 4 of the 31 models tracked here at a comfortable speed, and another 15 slowly. No chat model fits comfortably, so this machine is best kept for speech tools and tiny assistants.

What limits this machine

Without a graphics card everything runs on the processor. System RAM (32 GB here) holds the model and moves data at roughly 70 GB/s, so even small models produce only a few words a second. It works for small assistants and speech tools, and is not realistic for image or video generation.

What it is good for

Chat and coding

Best chat pick: Gemma 4 31B at about 2 words a second. Best coding pick: Qwen3 Coder 30B at about 8.

Images

Image models need more memory than this machine has.

Video

Video models need more memory than this machine has.

Voice and music

Fastest models here

No text model runs at a useful speed.

Memory headroom

Nothing fits comfortably, so there is no headroom to speak of.

What does not fit, and by how much

How it compares

Is an upgrade worth it?

The biggest single jump is PC with RTX 3090 (24 GB) and 32 GB RAM, which would unlock 19 more models, including Gemma 4 E2B, Qwen 3.5 4B, Gemma 4 E4B.

Questions

What is the best local AI model for CPU only 32GB?

For chat, Gemma 4 31B is the best fit, at about 2 words a second. For coding, Qwen3 Coder 30B.

How fast is CPU only 32GB for local AI?

Gemma 4 12B runs at about 4 words a second (slow); gpt-oss 20B runs at about 7 words a second (slow); Qwen 3.5 35B A3B does not fit.

Can CPU only 32GB run Qwen 3.5 35B A3B?

No. Qwen 3.5 35B A3B needs about 23.6 GB and this machine tops out near 22 GB.

The 19 models that run on this machine

ModelVerdictSpeedNeeds
ACE-Step 1.5Runs wellabout 25 min12 GB
Qwen3 TTS 1.7BRuns wellabout 8 min8 GB
Whisper large-v3 turboRuns wellabout 10 min4 GB
Kokoro TTSRuns wellabout 2 min2 GB
Qwen3 Coder 30BSlow8 words/s20.4 GB
Gemma 4 26B A4BSlow6 words/s19.4 GB
gpt-oss 20BSlow7 words/s15.2 GB
Phi-4 14BSlow3 words/s10.1 GB
Gemma 4 12BSlow4 words/s8.9 GB
Gemma 4 E4BSlow9 words/s7.4 GB
Qwen 3.5 9BSlow5 words/s7.4 GB
Qwen3 VL 8BSlow5 words/s6.9 GB
DeepSeek R1 8BSlow5 words/s6 GB
Qwen2.5 Coder 7BSlow6 words/s5.4 GB
Gemma 4 E2BSlow14 words/s5.3 GB
Qwen 3.5 4BSlow9 words/s4 GB
Gemma 4 31BVery slow2 words/s21.5 GB
DeepSeek R1 32BVery slow2 words/s21.5 GB
Qwen 3.6 27BVery slow2 words/s19.9 GB

What an upgrade would unlock

As an Amazon Associate I earn from qualifying purchases.

PC with RTX 3060 12GB (12 GB) and 32 GB RAM

Unlocks 11 more: Gemma 4 E2B, Qwen 3.5 4B, Gemma 4 E4B, Qwen 3.5 9B and more.

See prices

PC with RTX 5060 Ti 16GB (16 GB) and 32 GB RAM

Unlocks 12 more: Gemma 4 E2B, Qwen 3.5 4B, Gemma 4 E4B, Qwen 3.5 9B and more.

See prices

PC with RTX 3090 (24 GB) and 32 GB RAM

Unlocks 19 more: Gemma 4 E2B, Qwen 3.5 4B, Gemma 4 E4B, Qwen 3.5 9B and more.

See prices
Too slow at home? Rent a cloud GPU by the hour and run any of these models.Try RunPod