← AI Hashrate

Best GPUs for 13B-class local LLMs

We treat '13B-class' as open-weight models with 12–20B parameters — 7 in our catalog. The table ranks every catalog GPU that fully fits at least one of them at Q4 / 8K context, by average estimated decode tok/s across the models it fits. Fit means weights + KV cache + 1 GB runtime overhead within 95% of VRAM — not weights alone. Methodology.

#GPUVRAM (GB)Bandwidth (GB/s)MSRPFitsAvg Q4 tok/sBuy
1B200 192GB SXM1928,000$40,0007/7356.1 est.
2MI300X 192GB1925,300$15,0007/7235.9 est.
3H200 141GB SXM51414,800$35,0007/7213.7 est.
4H100 80GB SXM5803,350$30,0007/7149.1 est.
5Gaudi 3 128GB1283,276$12,5007/7145.8 est.
6A100 80GB SXM4802,039$15,0007/792.3 incl. measured
7RTX 5090 32GB321,792$1,9997/789.5 incl. measuredSearch Amazon (affiliate)
8RTX 4090 24GB241,008$1,5997/762.5 incl. measuredGIGABYTE GeForce RTX 4090 Gaming OC 24GB (affiliate) · More listings
9RX 7900 XTX 24GB24960$9997/745.7 incl. measuredASRock Radeon RX 7900 XTX Phantom Gaming 24GB OC (affiliate) · More listings
10RTX 3090 Ti 24GB241,008$1,9997/744.8 est.Search Amazon (affiliate)
11RTX 4090D 24GB (China)241,008$1,5997/744.8 est.
12RTX 3090 24GB24936$1,4997/743.7 incl. measuredMSI GeForce RTX 3090 Ventus 3X 24G OC (affiliate) · More listings
13RTX 5080 16GB16960$9997/742.8 est.NVIDIA GeForce RTX 5080 Founders Edition (affiliate) · More listings
14RTX 6000 Ada 48GB48960$6,8007/742.8 est.Search Amazon (affiliate)
15RTX 5070 Ti 16GB16896$7497/739.9 est.Search Amazon (affiliate)
16Mac Studio M3 Ultra 96GB96819$3,9997/738.8 incl. measuredSearch Amazon (affiliate)
17L40S 48GB48864$8,0007/738.4 est.Search Amazon (affiliate)
18Mac Studio M2 Ultra 192GB192800$6,9997/735.6 est.Search Amazon (affiliate)
19RTX 5070 12GB12672$5491/735.6 est.Search Amazon (affiliate)
20RX 7900 XT 20GB20800$8997/735.6 est.Search Amazon (affiliate)
21RTX A6000 48GB48768$4,5007/734.2 est.Search Amazon (affiliate)
22RTX 4070 Ti Super 16GB16672$7997/733.5 incl. measuredSearch Amazon (affiliate)
23RTX 4080 Super 16GB16736$9997/732.7 est.ASUS ProArt GeForce RTX 4080 Super OC (affiliate) · More listings
24RTX 4070 Ti 12GB12554$7991/729.4 est.Search Amazon (affiliate)
25RX 9070 16GB16640$5497/728.5 est.Search Amazon (affiliate)
26RX 9070 XT 16GB16640$5997/728.5 est.Search Amazon (affiliate)
27RX 7800 XT 16GB16624$4997/727.8 est.Search Amazon (affiliate)
28RTX 4070 12GB12504$5991/726.7 est.ASUS Dual GeForce RTX 4070 OC 12GB (affiliate) · More listings
29RTX 4070 Super 12GB12504$5991/726.7 est.Search Amazon (affiliate)
30RX 7900 GRE 16GB16576$5497/725.6 est.Search Amazon (affiliate)
31Arc A770 16GB16560$3497/724.9 est.Search Amazon (affiliate)
32MacBook Pro M4 Max 128GB128546$4,3997/724.3 est.Search Amazon (affiliate)
33Arc B580 12GB12456$2491/724.2 est.Search Amazon (affiliate)
34RX 7700 XT 12GB12432$4491/722.9 est.Search Amazon (affiliate)
35RTX 5060 Ti 16GB16448$4297/720.0 est.Search Amazon (affiliate)
36RTX 3060 12GB12360$3291/719.1 est.MSI GeForce RTX 3060 Ventus 2X 12G OC (affiliate) · More listings
37MacBook Pro M3 Max 128GB128400$4,1997/717.8 est.Search Amazon (affiliate)
38RTX 4060 Ti 16GB16288$4997/714.8 incl. measuredSearch Amazon (affiliate)
39T4 16GB16320$2,0007/714.3 est.Search Amazon (affiliate)
40RX 7600 XT 16GB16288$3297/712.8 est.Search Amazon (affiliate)
41Mac mini M4 Pro 64GB64273$2,1997/712.1 est.Search Amazon (affiliate)
42Mac mini M4 32GB32120$9997/75.4 est.Search Amazon (affiliate)

Averages mix estimated rows with measured anchors where they exist — per-model labels live on each hardware page. MSRP is launch list price, not live retail — Amazon prices change; click a Buy link for the current price. GPUs that fit none of this class are omitted.

More lists: 7B-class models · 30B-class models · 70B+ models · Under $500 · Under $1,000 · Under $2,000 · 8 GB VRAM · 12 GB VRAM · 16 GB VRAM · 24 GB VRAM · For Qwen3 · For DeepSeek · For Llama · For Gemma · For GLM · For coding · For reasoning · MoE picks