What can NVIDIA GeForce RTX 4090 run for agents?
Build: RTX 4090 + Ryzen 9 7950X + 64GB DDR5
Memory: 24 GB VRAM + 64 GB system RAM
Runner: llama.cpp / Ollama (CUDA)
Runs comfortably120 models
Ranked by fit for agents use case + predicted speed. Click a row for VRAM breakdown.
Quant: Q8_0Context: 8,192VRAM: 14.7 GBHeadroom: 9.3 GBTTFT: fastollama run hermes3:8b77tok/sEstimated
Quant: Q8_0Context: 8,192VRAM: 14.7 GBHeadroom: 9.3 GBTTFT: fast
ollama run hermes3:8b77
tok/s
Estimated
Weights
8.50 GB
KV cache
4.00 GB
Activations
0.43 GB
Runtime
1.80 GB
Time to first token (prefill, 512-token prompt): ~248 ms (fast)
Quant: Q5_K_MContext: 8,192VRAM: 19.8 GBHeadroom: 4.2 GBTTFT: fastollama run qwen2.5:14b68tok/sEstimated
Quant: Q5_K_MContext: 8,192VRAM: 19.8 GBHeadroom: 4.2 GBTTFT: fast
ollama run qwen2.5:14b68
tok/s
Estimated
Weights
10.50 GB
KV cache
7.00 GB
Activations
0.53 GB
Runtime
1.80 GB
Time to first token (prefill, 512-token prompt): ~434 ms (fast)
Quant: Q4_K_MContext: 2,048VRAM: 19.5 GBHeadroom: 4.5 GBTTFT: noticeableollama run dolphin-mistral:24b45tok/sEstimated
Quant: Q4_K_MContext: 2,048VRAM: 19.5 GBHeadroom: 4.5 GBTTFT: noticeable
ollama run dolphin-mistral:24b45
tok/s
Estimated
Weights
14.00 GB
KV cache
3.00 GB
Activations
0.70 GB
Runtime
1.80 GB
Time to first token (prefill, 512-token prompt): ~744 ms (noticeable)
Quant: Q8_0Context: 8,192VRAM: 10.8 GBHeadroom: 13.2 GBTTFT: fastollama run qwen2.5:7b88tok/sEstimated
Quant: Q8_0Context: 8,192VRAM: 10.8 GBHeadroom: 13.2 GBTTFT: fast
ollama run qwen2.5:7b88
tok/s
Estimated
Weights
8.10 GB
KV cache
0.47 GB
Activations
0.41 GB
Runtime
1.80 GB
Time to first token (prefill, 512-token prompt): ~217 ms (fast)
Quant: Q4_K_MContext: 8,192VRAM: 12.2 GBHeadroom: 11.8 GBTTFT: fastollama run ornith:9b121tok/sEstimated
Quant: Q4_K_MContext: 8,192VRAM: 12.2 GBHeadroom: 11.8 GBTTFT: fast
ollama run ornith:9b121
tok/s
Estimated
Weights
5.60 GB
KV cache
4.50 GB
Activations
0.29 GB
Runtime
1.80 GB
Time to first token (prefill, 512-token prompt): ~279 ms (fast)
Quant: Q4_K_MContext: 8,192VRAM: 11.0 GBHeadroom: 13.0 GBTTFT: fast136tok/sEstimated
Quant: Q4_K_MContext: 8,192VRAM: 11.0 GBHeadroom: 13.0 GBTTFT: fast
136
tok/s
Estimated
Weights
4.90 GB
KV cache
4.00 GB
Activations
0.25 GB
Runtime
1.80 GB
Time to first token (prefill, 512-token prompt): ~248 ms (fast)
Quant: Q5_K_MContext: 8,192VRAM: 10.7 GBHeadroom: 13.3 GBTTFT: fastollama run mistral:7b136tok/sEstimated
Quant: Q5_K_MContext: 8,192VRAM: 10.7 GBHeadroom: 13.3 GBTTFT: fast
ollama run mistral:7b136
tok/s
Estimated
Weights
5.10 GB
KV cache
3.50 GB
Activations
0.26 GB
Runtime
1.80 GB
Time to first token (prefill, 512-token prompt): ~217 ms (fast)
Quant: Q4_K_MContext: 8,192VRAM: 16.3 GBHeadroom: 7.7 GBTTFT: fastollama run hf.co/JetBrains/Mellum2-12B-A2.5B-Thinking-GGUF-Q4_K_M89tok/sEstimated
Quant: Q4_K_MContext: 8,192VRAM: 16.3 GBHeadroom: 7.7 GBTTFT: fast
ollama run hf.co/JetBrains/Mellum2-12B-A2.5B-Thinking-GGUF-Q4_K_M89
tok/s
Estimated
Weights
8.00 GB
KV cache
6.08 GB
Activations
0.41 GB
Runtime
1.80 GB
Time to first token (prefill, 512-token prompt): ~377 ms (fast)
Quant: Q4_K_MContext: 2,048VRAM: 19.5 GBHeadroom: 4.5 GBTTFT: noticeableollama run mistral-small:24b45tok/sEstimated
Quant: Q4_K_MContext: 2,048VRAM: 19.5 GBHeadroom: 4.5 GBTTFT: noticeable
ollama run mistral-small:24b45
tok/s
Estimated
Weights
14.00 GB
KV cache
3.00 GB
Activations
0.70 GB
Runtime
1.80 GB
Time to first token (prefill, 512-token prompt): ~744 ms (noticeable)
Quant: FP16Context: 8,192VRAM: 19.8 GBHeadroom: 4.2 GBTTFT: fastollama run llama3.1:8b41tok/sEstimated
Quant: FP16Context: 8,192VRAM: 19.8 GBHeadroom: 4.2 GBTTFT: fast
ollama run llama3.1:8b41
tok/s
Estimated
Weights
16.10 GB
KV cache
1.07 GB
Activations
0.81 GB
Runtime
1.80 GB
Time to first token (prefill, 512-token prompt): ~248 ms (fast)
Quant: Q4_K_MContext: 8,192VRAM: 11.1 GBHeadroom: 12.9 GBTTFT: fastollama run dolphin3:8b136tok/sEstimated
Quant: Q4_K_MContext: 8,192VRAM: 11.1 GBHeadroom: 12.9 GBTTFT: fast
ollama run dolphin3:8b136
tok/s
Estimated
Weights
5.00 GB
KV cache
4.00 GB
Activations
0.26 GB
Runtime
1.80 GB
Time to first token (prefill, 512-token prompt): ~248 ms (fast)
Quant: Q4_K_MContext: 8,192VRAM: 17.6 GBHeadroom: 6.4 GBTTFT: fastollama run qwen3:14b78tok/sEstimated
Quant: Q4_K_MContext: 8,192VRAM: 17.6 GBHeadroom: 6.4 GBTTFT: fast
ollama run qwen3:14b78
tok/s
Estimated
Weights
8.40 GB
KV cache
7.00 GB
Activations
0.43 GB
Runtime
1.80 GB
Time to first token (prefill, 512-token prompt): ~434 ms (fast)
Runs with tradeoffs73 models
Tight VRAM, partial CPU offload, or context-limited.
Quant: Q4_K_MContext: 8,192VRAM: 23.9 GBHeadroom: 0.1 GBTTFT: noticeable- • Tight VRAM fit — only 0.1 GB headroom left for context growth
ollama run qwen2.5-coder:32b34tok/sEstimated
Quant: Q4_K_MContext: 8,192VRAM: 23.9 GBHeadroom: 0.1 GBTTFT: noticeable
- • Tight VRAM fit — only 0.1 GB headroom left for context growth
ollama run qwen2.5-coder:32b34
tok/s
Estimated
Weights
19.00 GB
KV cache
2.15 GB
Activations
0.96 GB
Runtime
1.80 GB
Time to first token (prefill, 512-token prompt): ~992 ms (noticeable)
Quant: Q4_K_MContext: 2,048VRAM: 21.9 GBHeadroom: 2.1 GBTTFT: noticeable- • Tight VRAM fit — only 2.1 GB headroom left for context growth
ollama run gemma4:26b-moe42tok/sEstimated
Quant: Q4_K_MContext: 2,048VRAM: 21.9 GBHeadroom: 2.1 GBTTFT: noticeable
- • Tight VRAM fit — only 2.1 GB headroom left for context growth
ollama run gemma4:26b-moe42
tok/s
Estimated
Weights
16.00 GB
KV cache
3.25 GB
Activations
0.80 GB
Runtime
1.80 GB
Time to first token (prefill, 512-token prompt): ~806 ms (noticeable)
Quant: Q4_K_MContext: 8,192VRAM: 36.8 GBHeadroom: 25.6 GBTTFT: noticeable- • Partial CPU offload: ~35% of layers run on CPU
- • CPU is the bottleneck — upgrading RAM bandwidth helps more than VRAM here
ollama run qwen3-coder:30b5tok/sEstimated
Quant: Q4_K_MContext: 8,192VRAM: 36.8 GBHeadroom: 25.6 GBTTFT: noticeable
- • Partial CPU offload: ~35% of layers run on CPU
- • CPU is the bottleneck — upgrading RAM bandwidth helps more than VRAM here
ollama run qwen3-coder:30b5
tok/s
Estimated
Weights
19.00 GB
KV cache
15.00 GB
Activations
0.96 GB
Runtime
1.80 GB
Time to first token (prefill, 512-token prompt): ~930 ms (noticeable)
Quant: Q4_K_MContext: 8,192VRAM: 36.8 GBHeadroom: 25.6 GBTTFT: noticeable- • Partial CPU offload: ~35% of layers run on CPU
- • CPU is the bottleneck — upgrading RAM bandwidth helps more than VRAM here
ollama run north-mini-code-1.05tok/sEstimated
Quant: Q4_K_MContext: 8,192VRAM: 36.8 GBHeadroom: 25.6 GBTTFT: noticeable
- • Partial CPU offload: ~35% of layers run on CPU
- • CPU is the bottleneck — upgrading RAM bandwidth helps more than VRAM here
ollama run north-mini-code-1.05
tok/s
Estimated
Weights
19.00 GB
KV cache
15.00 GB
Activations
0.96 GB
Runtime
1.80 GB
Time to first token (prefill, 512-token prompt): ~930 ms (noticeable)
Quant: Q4_K_MContext: 8,192VRAM: 37.3 GBHeadroom: 25.1 GBTTFT: noticeable- • Partial CPU offload: ~36% of layers run on CPU
- • CPU is the bottleneck — upgrading RAM bandwidth helps more than VRAM here
ollama run glm-4.7-flash5tok/sEstimated
Quant: Q4_K_MContext: 8,192VRAM: 37.3 GBHeadroom: 25.1 GBTTFT: noticeable
- • Partial CPU offload: ~36% of layers run on CPU
- • CPU is the bottleneck — upgrading RAM bandwidth helps more than VRAM here
ollama run glm-4.7-flash5
tok/s
Estimated
Weights
19.00 GB
KV cache
15.50 GB
Activations
0.96 GB
Runtime
1.80 GB
Time to first token (prefill, 512-token prompt): ~961 ms (noticeable)
Quant: Q4_K_MContext: 8,192VRAM: 39.3 GBHeadroom: 23.1 GBTTFT: noticeable- • Partial CPU offload: ~39% of layers run on CPU
- • CPU is the bottleneck — upgrading RAM bandwidth helps more than VRAM here
ollama run laguna-xs-2.14tok/sEstimated
Quant: Q4_K_MContext: 8,192VRAM: 39.3 GBHeadroom: 23.1 GBTTFT: noticeable
- • Partial CPU offload: ~39% of layers run on CPU
- • CPU is the bottleneck — upgrading RAM bandwidth helps more than VRAM here
ollama run laguna-xs-2.14
tok/s
Estimated
Weights
20.00 GB
KV cache
16.50 GB
Activations
1.01 GB
Runtime
1.80 GB
Time to first token (prefill, 512-token prompt): ~1023 ms (noticeable)
Quant: Q4_K_MContext: 8,192VRAM: 41.4 GBHeadroom: 21.0 GBTTFT: noticeable- • Partial CPU offload: ~42% of layers run on CPU
- • CPU is the bottleneck — upgrading RAM bandwidth helps more than VRAM here
ollama run ornith:35b4tok/sEstimated
Quant: Q4_K_MContext: 8,192VRAM: 41.4 GBHeadroom: 21.0 GBTTFT: noticeable
- • Partial CPU offload: ~42% of layers run on CPU
- • CPU is the bottleneck — upgrading RAM bandwidth helps more than VRAM here
ollama run ornith:35b4
tok/s
Estimated
Weights
21.00 GB
KV cache
17.50 GB
Activations
1.06 GB
Runtime
1.80 GB
Time to first token (prefill, 512-token prompt): ~1085 ms (noticeable)
Quant: Q8_0Context: 8,192VRAM: 50.4 GBHeadroom: 12.0 GBTTFT: noticeable- • Partial CPU offload: ~52% of layers run on CPU
- • CPU is the bottleneck — upgrading RAM bandwidth helps more than VRAM here
ollama run qwen3:30b2tok/sEstimated
Quant: Q8_0Context: 8,192VRAM: 50.4 GBHeadroom: 12.0 GBTTFT: noticeable
- • Partial CPU offload: ~52% of layers run on CPU
- • CPU is the bottleneck — upgrading RAM bandwidth helps more than VRAM here
ollama run qwen3:30b2
tok/s
Estimated
Weights
32.00 GB
KV cache
15.00 GB
Activations
1.61 GB
Runtime
1.80 GB
Time to first token (prefill, 512-token prompt): ~930 ms (noticeable)
What if you upgraded?
Hypothetical scenarios. We re-ran the compatibility engine for each.
+32 GB system RAM
~$80–150
Doubles your CPU-offload working set. Helps when models don't quite fit in VRAM.
Unlocks: 64 new comfortable, 79 new tradeoff
- • Qwen 3 0.6B
- • Llama 3.2 3B Instruct
- • Qwen 3 1.7B
- • Gemma 3 270M
Upgrade to NVIDIA RTX PRO 4500 Blackwell
see current pricing
32 GB VRAM (vs your 24 GB) plus a bandwidth jump from ~1008 GB/s to ~896 GB/s.
Unlocks: 103 new comfortable
- • Qwen 3 0.6B
- • Qwen 3 30B-A3B
- • Qwen 2.5 Coder 32B Instruct
- • Qwen3.6 27B
Add a second NVIDIA GeForce RTX 4090
~$1899
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: 116 new comfortable
- • Qwen 3 0.6B
- • Qwen 3 30B-A3B
- • Qwen 2.5 Coder 32B Instruct
- • Qwen3.6 27B
Some links above are affiliate links. We may earn a commission at no extra cost to you. How we make money.
Won't runtop 5 popular models
Need more memory than you have. Shown for orientation.
Even with CPU offload, needs more memory than your VRAM (24 GB) + 60% of system RAM (38 GB) combined.
—
Even with CPU offload, needs more memory than your VRAM (24 GB) + 60% of system RAM (38 GB) combined.
—
Even with CPU offload, needs more memory than your VRAM (24 GB) + 60% of system RAM (38 GB) combined.
—
Even with CPU offload, needs more memory than your VRAM (24 GB) + 60% of system RAM (38 GB) combined.
—
Even with CPU offload, needs more memory than your VRAM (24 GB) + 60% of system RAM (38 GB) combined.
—
Even with CPU offload, needs more memory than your VRAM (24 GB) + 60% of system RAM (38 GB) combined.
—
Even with CPU offload, needs more memory than your VRAM (24 GB) + 60% of system RAM (38 GB) combined.
—
Even with CPU offload, needs more memory than your VRAM (24 GB) + 60% of system RAM (38 GB) combined.
—
Even with CPU offload, needs more memory than your VRAM (24 GB) + 60% of system RAM (38 GB) combined.
—
Even with CPU offload, needs more memory than your VRAM (24 GB) + 60% of system RAM (38 GB) combined.
—
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.Want a specific benchmark we don't have? Email Contact support and we'll prioritize it.