What can AMD Radeon RX 7900 XTX run for chat?

Build: AMD Radeon RX 7900 XTX + — + 32 GB RAM (windows)

Memory: 24 GB VRAM + 32 GB system RAM
Runner: llama.cpp (Vulkan)
AnyChatCodingAgentsReasoningVisionLong contextCreative

Runs comfortably
107 models

Ranked by fit for chat use case + predicted speed. Click a row for VRAM breakdown.

#1Qwen 3 0.6B
0.6B
qwen
Commercial OK
Quant: Q4_K_MContext: 8,192VRAM: 1.6 GBHeadroom: 22.4 GBTTFT: instant
1193
tok/s
Estimated
Weights
0.30 GB
KV cache
0.30 GB
Activations
0.02 GB
Runtime
1.00 GB
Time to first token (prefill, 512-token prompt): ~13 ms (instant)
Model details →
#2Llama 3.2 3B Instruct
3B
llama
Commercial OK
Quant: Q8_0Context: 8,192VRAM: 6.1 GBHeadroom: 17.9 GBTTFT: instant
ollama run llama3.2:3b
136
tok/s
Estimated
Weights
3.40 GB
KV cache
1.50 GB
Activations
0.18 GB
Runtime
1.00 GB
Time to first token (prefill, 512-token prompt): ~63 ms (instant)
Model details →
#3Qwen 3 1.7B
1.7B
qwen
Commercial OK
Quant: Q4_K_MContext: 8,192VRAM: 2.8 GBHeadroom: 21.2 GBTTFT: instant
421
tok/s
Estimated
Weights
0.90 GB
KV cache
0.85 GB
Activations
0.05 GB
Runtime
1.00 GB
Time to first token (prefill, 512-token prompt): ~35 ms (instant)
Model details →
#4TinyLlama 1.1B Chat v1.0
1.1B
llama
Commercial OK
Quant: Q4_K_MContext: 2,048VRAM: 1.8 GBHeadroom: 22.2 GBTTFT: instant
650
tok/s
Estimated
Weights
0.60 GB
KV cache
0.14 GB
Activations
0.03 GB
Runtime
1.00 GB
Time to first token (prefill, 512-token prompt): ~23 ms (instant)
Model details →
#5Gemma 2 2B Instruct
2B
gemma
Commercial OK
Quant: Q4_K_MContext: 8,192VRAM: 3.2 GBHeadroom: 20.8 GBTTFT: instant
358
tok/s
Estimated
Weights
1.10 GB
KV cache
1.00 GB
Activations
0.06 GB
Runtime
1.00 GB
Time to first token (prefill, 512-token prompt): ~42 ms (instant)
Model details →
#6DeepSeek V2 Lite Chat
15.7B
deepseek
Commercial OK
Quant: Q4_K_MContext: 8,192VRAM: 17.9 GBHeadroom: 6.1 GBTTFT: instant
298
tok/s
Estimated
Weights
8.60 GB
KV cache
7.85 GB
Activations
0.44 GB
Runtime
1.00 GB
Time to first token (prefill, 512-token prompt): ~50 ms (instant)
Model details →
#7Kumru 2B
2.4B
mistral
Commercial OK
Quant: Q4_K_MContext: 8,192VRAM: 3.8 GBHeadroom: 20.2 GBTTFT: instant
ollama run alibayram/kumru:latest
298
tok/s
Estimated
Weights
1.50 GB
KV cache
1.20 GB
Activations
0.08 GB
Runtime
1.00 GB
Time to first token (prefill, 512-token prompt): ~50 ms (instant)
Model details →
#8Falcon 3 3B Instruct
3B
falcon
Commercial OK
Quant: Q4_K_MContext: 8,192VRAM: 4.3 GBHeadroom: 19.7 GBTTFT: instant
239
tok/s
Estimated
Weights
1.70 GB
KV cache
1.50 GB
Activations
0.09 GB
Runtime
1.00 GB
Time to first token (prefill, 512-token prompt): ~63 ms (instant)
Model details →
#9Dolphin 3.0 Llama 3.2 3B
3B
dolphin
Commercial OK
Quant: Q4_K_MContext: 8,192VRAM: 4.4 GBHeadroom: 19.6 GBTTFT: instant
239
tok/s
Estimated
Weights
1.80 GB
KV cache
1.50 GB
Activations
0.10 GB
Runtime
1.00 GB
Time to first token (prefill, 512-token prompt): ~63 ms (instant)
Model details →
#10Hermes 3 Llama 3.2 3B
3B
hermes
Commercial OK
Quant: Q4_K_MContext: 8,192VRAM: 4.4 GBHeadroom: 19.6 GBTTFT: instant
239
tok/s
Estimated
Weights
1.80 GB
KV cache
1.50 GB
Activations
0.10 GB
Runtime
1.00 GB
Time to first token (prefill, 512-token prompt): ~63 ms (instant)
Model details →
#11Granite 3.0 2B Instruct
2B
granite
Commercial OK
Quant: Q4_K_MContext: 4,096VRAM: 3.1 GBHeadroom: 20.9 GBTTFT: instant
358
tok/s
Estimated
Weights
1.50 GB
KV cache
0.50 GB
Activations
0.08 GB
Runtime
1.00 GB
Time to first token (prefill, 512-token prompt): ~42 ms (instant)
Model details →
#12Mistral Turkish v2 (brooqs)
7.2B
mistral
Quant: Q4_0Context: 8,192VRAM: 8.9 GBHeadroom: 15.1 GBTTFT: fast
ollama run brooqs/mistral-turkish-v2:latest
107
tok/s
Estimated
Weights
4.10 GB
KV cache
3.60 GB
Activations
0.21 GB
Runtime
1.00 GB
Time to first token (prefill, 512-token prompt): ~150 ms (fast)
Model details →

Runs with tradeoffs
24 models

Tight VRAM, partial CPU offload, or context-limited.

Mistral Nemo 12B Instruct
12B
mistral
Commercial OK
Quant: Q8_0Context: 8,192VRAM: 20.7 GBHeadroom: 3.3 GBTTFT: fast
  • • Tight VRAM fit — only 3.3 GB headroom left for context growth
ollama run mistral-nemo:12b
34
tok/s
Estimated
Weights
13.00 GB
KV cache
6.00 GB
Activations
0.66 GB
Runtime
1.00 GB
Time to first token (prefill, 512-token prompt): ~250 ms (fast)
Model details →
Gemma 3 12B
12B
gemma
Commercial OK
Quant: Q8_0Context: 8,192VRAM: 20.7 GBHeadroom: 3.3 GBTTFT: fast
  • • Tight VRAM fit — only 3.3 GB headroom left for context growth
ollama run gemma3:12b
34
tok/s
Estimated
Weights
13.00 GB
KV cache
6.00 GB
Activations
0.66 GB
Runtime
1.00 GB
Time to first token (prefill, 512-token prompt): ~250 ms (fast)
Model details →
Gemma 3 27B
27B
gemma
Commercial OK
Quant: Q4_K_MContext: 2,048VRAM: 21.2 GBHeadroom: 2.8 GBTTFT: noticeable
  • • Tight VRAM fit — only 2.8 GB headroom left for context growth
ollama run gemma3:27b
27
tok/s
Estimated
Weights
16.00 GB
KV cache
3.38 GB
Activations
0.80 GB
Runtime
1.00 GB
Time to first token (prefill, 512-token prompt): ~563 ms (noticeable)
Model details →
Gemma 4 31B Dense
31B
gemma
Commercial OK
Quant: Q4_K_MContext: 2,048VRAM: 23.8 GBHeadroom: 0.2 GBTTFT: noticeable
  • • Tight VRAM fit — only 0.2 GB headroom left for context growth
ollama run gemma4:31b
23
tok/s
Estimated
Weights
18.00 GB
KV cache
3.88 GB
Activations
0.90 GB
Runtime
1.00 GB
Time to first token (prefill, 512-token prompt): ~646 ms (noticeable)
Model details →
Yi 1.5 34B
34B
yi
Commercial OK
Quant: Q4_K_MContext: 8,192VRAM: 39.0 GBHeadroom: 4.2 GBTTFT: noticeable
  • • Partial CPU offload: ~38% of layers run on CPU
ollama run yi:34b
21
tok/s
Estimated
Weights
20.00 GB
KV cache
17.00 GB
Activations
1.01 GB
Runtime
1.00 GB
Time to first token (prefill, 512-token prompt): ~709 ms (noticeable)
Model details →
Mixtral 8x7B Instruct
47B
mixtral
Commercial OK
Quant: Q4_K_MContext: 2,048VRAM: 36.3 GBHeadroom: 6.9 GBTTFT: noticeable
  • • Partial CPU offload: ~34% of layers run on CPU
ollama run mixtral:8x7b
15
tok/s
Estimated
Weights
28.00 GB
KV cache
5.88 GB
Activations
1.40 GB
Runtime
1.00 GB
Time to first token (prefill, 512-token prompt): ~980 ms (noticeable)
Model details →
Jamba 1.5 Mini
52B
other
Commercial OK
Quant: Q4_K_MContext: 2,048VRAM: 39.0 GBHeadroom: 4.2 GBTTFT: fast
  • • Partial CPU offload: ~38% of layers run on CPU
60
tok/s
Estimated
Weights
30.00 GB
KV cache
6.50 GB
Activations
1.50 GB
Runtime
1.00 GB
Time to first token (prefill, 512-token prompt): ~250 ms (fast)
Model details →
Pixtral 12B
12B
mistral
Commercial OK
Quant: Q8_0Context: 8,192VRAM: 20.7 GBHeadroom: 3.3 GBTTFT: fast
  • • Tight VRAM fit — only 3.3 GB headroom left for context growth
ollama run pixtral:12b
34
tok/s
Estimated
Weights
13.00 GB
KV cache
6.00 GB
Activations
0.66 GB
Runtime
1.00 GB
Time to first token (prefill, 512-token prompt): ~250 ms (fast)
Model details →

What if you upgraded?

Hypothetical scenarios. We re-ran the compatibility engine for each.

+32 GB system RAM

Check the current price

Adds 32 GB to your CPU-offload working set. Helps when models don't quite fit in VRAM.

Unlocks: 5 new comfortable, 36 new tradeoff

  • • Gemma 3 270M
  • • SmolLM2 135M Instruct
  • • SmolLM2 360M Instruct
  • • VBART Large (Turkish Summarization)
Shop this upgrade↗

Upgrade to AMD Instinct MI210

Launch MSRP $8,500 (2022). Check the current price.

64 GB VRAM (vs your 24 GB) plus a bandwidth jump from ~960 GB/s to ~? GB/s.

Unlocks: 39 new comfortable

  • • Qwen 3 30B-A3B
  • • Llama 3.3 70B Instruct
  • • Qwen 2.5 Coder 32B Instruct
  • • Qwen 3 32B
Shop this upgrade↗

Add a second AMD Radeon RX 7900 XTX

Launch MSRP $999 (2022). Check the current price.

Tensor parallelism splits the model across both cards, effectively doubling VRAM. Bandwidth doesn't double — runs ~1.5× the single-card speed in practice.

Unlocks: 27 new comfortable

  • • Qwen 3 30B-A3B
  • • Qwen 2.5 Coder 32B Instruct
  • • Qwen 3 32B
  • • Gemma 4 31B Dense
Shop this upgrade↗

Some links above are affiliate links. We may earn a commission at no extra cost to you. How we make money.

Won't run
top 5 popular models

Need more memory than you have. Shown for orientation.

DeepSeek V4 Pro (1.6T MoE)
1600B
deepseek
Commercial OK

Even with CPU offload, needs more memory than your VRAM (24 GB) + 60% of system RAM (19 GB) combined.

—
Qwen 3 235B-A22B
235B
qwen
Commercial OK

Even with CPU offload, needs more memory than your VRAM (24 GB) + 60% of system RAM (19 GB) combined.

—
DeepSeek R1 (671B reasoning)
671B
deepseek
Commercial OK

Even with CPU offload, needs more memory than your VRAM (24 GB) + 60% of system RAM (19 GB) combined.

—
Llama 4 Scout
109B
llama
Commercial OK

Even with CPU offload, needs more memory than your VRAM (24 GB) + 60% of system RAM (19 GB) combined.

—
DeepSeek V4 Flash (284B MoE)
284B
deepseek
Commercial OK

Even with CPU offload, needs more memory than your VRAM (24 GB) + 60% of system RAM (19 GB) combined.

—
Cloud GPU rental

Not enough VRAM for these models? Rent a GPU by the hour

The models above need more memory than 24 GB VRAM + 32 GB system RAM. You can rent a GPU with enough VRAM by the hour to run one of them, or to test it before you buy hardware. For daily use, owning a card usually costs less over time.

RunPod

billed per GPU-hour

Hourly GPU pods, in community-cloud and secure-cloud tiers.

The two tiers run on different hosts and terms. Check which tier a pod is on before a long job.

Rent on RunPod

Vast.ai

billed per GPU-hour

Marketplace of host-listed GPUs; price and reliability vary by host.

Host quality varies. Filter for verified hosts and check the host's reliability before a long run.

Rent on Vast.ai

Referral links: RunLocalAI may earn a commission or account credits if you sign up and rent, at no extra cost to you. Hourly rates and which GPUs are listed change all the time, so check the provider's page before you start. Editorial opinions are independent of the referral relationship; see how we make money.

How to read these numbers

Measured here
Measured here - RunLocalAI ran this exact combo on owner hardware with public evidence.

Source-backed
Source-backed / community - a reproduced public source supports the speed, but it is not labeled as owner-measured.

Extrapolated
Extrapolated - predicted from a measured benchmark on similar-bandwidth hardware.

Estimated
Estimated - formula based on VRAM bandwidth and model architecture; not a benchmark row.

RunLocalAI Will-It-Run Framework →

Want a specific benchmark we don't have? Email Contact support and we'll prioritize it.