What AI can an RTX 5090 run?

Quick verdict

PC with RTX 5090 (32 GB) and 32 GB RAM can run 26 of the 31 models tracked here at a comfortable speed, and another 2 slowly. The largest chat or coding model that fits is Qwen 3.6 35B A3B, which needs about 24.6 GB and generates roughly 190+ words a second. The quickest is Gemma 4 E2B at about 190+ words a second, which feels faster than you can read.

What limits this machine

Video memory (VRAM) is the limit: 32 GB on this card. A model must fit in VRAM to run fast. Anything larger spills into system RAM (32 GB in this setup), which works but is slow, because system RAM moves data at roughly 70 GB/s while the card moves 1792 GB/s.

What it is good for

Chat and coding

Best chat pick: Qwen 3.5 35B A3B at about 190+ words a second. Best coding pick: Qwen 3.6 35B A3B at about 190+.

Images

• Stable Diffusion XL: runs great, about about 4 s per image
• FLUX.2 klein 4B: runs great, about about 6 s per image
• Z-Image Turbo 6B: runs great, about about 8 s per image

Video

• LTX-2.3: runs great, about about 40 s per 4 second clip
• Wan 2.2 5B: runs great, about about 2 min per 4 second clip
• Wan 2.2 14B: slow, about about 8 min per 4 second clip

Voice and music

• Whisper large-v3 turbo: runs great, about about 15 s per 10 minutes of audio
• Kokoro TTS: runs great, about about 2 s per minute of speech
• Qwen3 TTS 1.7B: runs great, about about 10 s per minute of speech
• ACE-Step 1.5: runs great, about about 25 s per 1 minute song

Fastest models here

• Gemma 4 E2B: about 190+ words a second, faster than you can read
• Qwen 3.5 4B: about 190+ words a second, faster than you can read
• Gemma 4 E4B: about 190+ words a second, faster than you can read
• Qwen 3.5 35B A3B: about 190+ words a second, faster than you can read
• Qwen3 Coder 30B: about 190+ words a second, faster than you can read

Memory headroom

Spare memory after loading the model decides how long a conversation or document can be before things slow down. With about 50 GB usable here:

• Qwen 3.6 35B A3B needs 24.6 GB, leaving 25 GB spare, plenty for long documents.
• Qwen 3.5 35B A3B needs 23.6 GB, leaving 26 GB spare, plenty for long documents.
• Gemma 4 31B needs 21.5 GB, leaving 28.1 GB spare, plenty for long documents.
• DeepSeek R1 32B needs 21.5 GB, leaving 28.1 GB spare, plenty for long documents.
• Qwen3 Coder 30B needs 20.4 GB, leaving 29.2 GB spare, plenty for long documents.

What does not fit, and by how much

• gpt-oss 120B needs about 68.8 GB and this machine tops out near 50 GB, so it is about 19 GB short.
• Qwen 3.5 122B A10B needs about 85.5 GB and this machine tops out near 50 GB, so it is about 36 GB short.
• DeepSeek V3 671B needs about 424.7 GB and this machine tops out near 50 GB, so it is about 375 GB short.

How it compares

• PC with RTX 3090, RTX 4090 or RTX 5090 Laptop and 32 GB RAM: 3 fewer models; drops Qwen 3.5 35B A3B, Qwen 3.6 35B A3B, LTX-2.3.
• PC with Radeon RX 7900 XTX (24 GB) and 32 GB RAM: 3 fewer models; drops Qwen 3.5 35B A3B, Qwen 3.6 35B A3B, LTX-2.3.

Is an upgrade worth it?

The biggest single jump is Ryzen AI Max+ 395 (Strix Halo), 128 GB, which would unlock 3 more models, including Qwen 3.5 122B A10B, gpt-oss 120B, Wan 2.2 14B.

Questions