Hardware

Radeon RX 6800

Radeon RX 6800 has 16 GB GDDR6 and 512 GB/s. See which open models fit, at which quant, and roughly how fast. Figures below use the default memory (16 GB), 32 GB system RAM, and DDR5-5600 for anything that spills. Updated 2026-10-24.

Memory16 GB GDDR6
Bandwidth512 GB/s
Kindgpu
Year2020

16GB GDDR6, 256-bit, 512 GB/s. Source.

RSS for new models on this device

Models that fit

ModelVerdictQuantContexttok/s
Qwen3.8 27B
27.3B
Tight fitUD-IQ4_XS8K19–26
Devstral Small 2 24B Instruct 2512
23.6B
Tight fitQ4_K_M4K19–26
Mistral Small 3.2 24B Instruct 2506
23.6B
Tight fitQ4_K_M4K19–26
Magistral Small 2509
23.6B
Tight fitQ4_K_M4K19–26
Swift Qwen3.8 27B
27.3B
Tight fitQ3_K_L8K19–26
gpt oss 20b
20.9B
Tight fitUD-Q8_K_XL8K70–95
gemma 3 27b it
27.0B
Tight fitUD-Q3_K_XL8K16–21
medgemma 27b text it
27.0B
Tight fitUD-Q3_K_XL8K16–21
Qwen3.6 27B
26.9B
Tight fitUD-Q3_K_XL4K19–26
Qwen3.5 27B
26.9B
Tight fitUD-Q3_K_XL4K19–26
GLM 4.7 Flash REAP 23B A3B
23.0B
Tight fitUD-Q4_K_XL8K122–167
GLM 4 32B 0414
32.6B
Tight fitQ3_K_S8K19–26
NousResearch Hermes 4 14B
14.8B
RunsQ5_K_M8K24–32
Qwen2.5 14B Instruct
14.8B
RunsQ5_K_M8K23–32
DeepSeek R1 Distill Qwen 14B
14.8B
Tight fitQ6_K8K21–28
Qwen2.5 Coder 14B Instruct 128K
14.8B
Tight fitQ6_K8K21–28
Qwen3 14B
14.8B
Tight fitUD-Q6_K_XL8K19–26
phi 4
14.7B
Tight fitQ6_K8K21–28
gemma 4 26B A4B it
25.2B
Tight fitUD-IQ4_XS8K68–92
gemma 4 31B it
30.7B
Tight fitQ3_K_S4K16–22
Qwen3 32B
32.8B
Tight fitUD-IQ3_XXS4K20–27
Qwen3 VL 32B Instruct
32.8B
Tight fitUD-IQ3_XXS4K20–27
bottlecapai ThinkingCap gemma 4 26B A4B it
25.2B
Tight fitIQ4_XS8K66–90
GLM 4.7 Flash
29.9B
Tight fitUD-Q3_K_XL8K84–114
Qwen3.5 35B A3B
34.7B
Tight fitUD-IQ3_S8K210–286
Qwen3.6 35B A3B
34.7B
Tight fitUD-IQ3_S8K208–284
Ministral 3 14B Instruct 2512
13.5B
Tight fitUD-Q6_K_XL8K21–29
Ministral 3 14B Reasoning 2512
13.5B
Tight fitUD-Q6_K_XL8K21–29
Qwen3 Coder 30B A3B Instruct
30.5B
Tight fitUD-Q3_K_XL8K130–178
Qwen3 30B A3B Instruct 2507
30.5B
Tight fitUD-Q3_K_XL8K130–177
Qwen3 VL 30B A3B Instruct
30.5B
Tight fitUD-Q3_K_XL8K130–177
Qwen3 30B A3B Thinking 2507
30.5B
Tight fitUD-Q3_K_XL8K130–177
Qwen2.5 Coder 32B Instruct
32.8B
Tight fitIQ3_XXS8K19–26
gemma 4 12b it
11.9B
Tight fitUD-Q8_K_XL8K17–23
Tiel Coder 35B A3B
34.7B
Tight fitUD-IQ3_XXS8K215–293
Qwen3.6 14B A3B FableVibes
13.8B
Tight fitQ8_08K84–114
LFM2.5 24B A2B
24.0B
RunsUD-Q3_K_XL8K178–242
Kwaipilot KAT Coder V2.5 Dev
34.7B
Tight fitIQ3_XXS4K101–138
Qwen2.5 32B Instruct
32.8B
RunsIQ2_S8K22–31
GLM 4.6V Flash
9.4B
RunsUD-Q8_K_XL8K23–31

105 more models in the catalog do not fit this default profile, or ranked below the first 40 that do.

Example: Tight fit fully on your Radeon RX 6800 at UD-IQ4_XS, 8K context, about 19–26 tok/s. Weights 13.3 GB plus 0.57 GB of KV cache — 14.3 GB of 14.7 GB usable. Weights shown in the verdict use GiB (the same 1024³ counting as GPU memory).