Your rig

RTX 2000 Ada · 16 GB

256 GB system RAM · Ada Lovelace · 224 GB/s · 14.4 GB usable VRAM
Edit hardware Export as JSON

Context length

KV cache at f16, 128K tokens

Quality floor
Q8_0 — near-lossless Q6_K Q5_K_M Q4_K_M — recommended Q3_K_M Q2_K — damaged
KV cache type
System RAM (GB)
Runtime
Ollamallama.cppLM StudiovLLM

4 of 5 runtimes drive this device.

Allow CPU offload Include partial-offload fits

Offloaded layers run at 60 GB/s host bandwidth — expect single-digit tokens per second.

Which one should I actually run?

Best models for your RTX 2000 Ada

Updated 8 Oct 2026 1 new this month

What do you want to do? Each row is the best-ranked model on that use case's shortlist that fits at Q4_K_M and 128K. The ordering is editorial; the sizes and speeds are computed. How the shortlists work.

RecommendationModelWhyQuant · totalTok/s est.
Best overall Qwen3.5 9B
Alibaba · Apache 2.0
The 8–12 GB default of 2026 — beats every older 8B, with vision. Q4_K_M · 10.0 GB 25 Runs great
Best for coding Qwen3.5 9B
Alibaba · Apache 2.0
The best coding model for 8–12 GB cards. Q4_K_M · 10.0 GB 25 Runs great
Best reasoning Qwen3.5 9B
Alibaba · Apache 2.0
Thinking mode on a 12 GB card — ahead of the 2025 R1 distills. Q4_K_M · 10.0 GB 25 Runs great
Best for writing Gemma 4 E4B
Google · Apache 2.0
Readable prose on a laptop. Q4_K_M · 6.9 GB 30 Runs great
Best vision Qwen3.5 9B
Alibaba · Apache 2.0
Vision on an 8–12 GB card. Q4_K_M · 10.0 GB 25 Runs great
Best for agents Fara 7B
Microsoft · MIT
Web computer-use only — clicks and fills forms from screenshots, and stops to ask before anything irreversible. Q4_K_M · 12.1 GB 29 Runs great
Best translation Qwen3.5 9B
Alibaba · Apache 2.0
Strong CJK and European translation in 8–12 GB. Q4_K_M · 10.0 GB 25 Runs great
Fastest good model LFM2.5 8B-A1B
Liquid AI · LFM Open License v1.0
1.5B active; designed for laptops without a GPU. Q4_K_M · 6.8 GB 62 Runs great
Best long context Qwen3.5 9B
Alibaba · Apache 2.0
262K native in 8–12 GB. Q4_K_M · 10.0 GB
at 128K context
25 Runs great

Everything that fits

Runs great

21 of 80 · fits with headroom for a longer prompt
ModelParamsQuantWeights +KV 128KTotal VRAM headroom Tok/sRuntime
Phi-4 14B 14.7B Q4_K_M 8.3 GB 3.13 GB 12.0 GB
2.4 GB free
16 Ollama Run
Qwen3.5 9B 9.65B Q4_K_M 5.4 GB 4.00 GB 10.0 GB
4.4 GB free
25 Ollama Run
Ornith 1.5 9B 9.41B Q4_K_M 5.3 GB 4.00 GB 9.9 GB
4.5 GB free
26 Ollama Run
LFM2.5 8B-A1B 8.47B MoE / 1.5B act Q4_K_M 4.8 GB 1.46 GB 6.8 GB
7.6 GB free
62 Ollama Run
Fara 7B 8.29B Q4_K_M 4.7 GB 6.84 GB 12.1 GB
2.3 GB free
29 llama.cpp Run
Gemma 4 E4B 8.0B Q4_K_M 4.5 GB 1.78 GB 6.9 GB
7.5 GB free
30 Ollama Run
Ling 3.0 Tiny 7.9B-A1.3B 7.9B MoE / 1.3B act Q4_K_M 4.4 GB 0.84 GB 5.9 GB
8.5 GB free
71 llama.cpp Run
DeepSeek-R1-Distill-Qwen 7B 7.62B Q4_K_M 4.3 GB 7.00 GB 11.9 GB
2.5 GB free
32 Ollama Run
Qwen2.5-Coder 7B 7.62B Q4_K_M 4.3 GB 7.00 GB 11.9 GB
2.5 GB free
32 Ollama Run
Mistral 7B Instruct v0.3 7.25B Q4_K_M 4.1 GB 4.00 GB 8.7 GB
5.7 GB free
33 Ollama Run
Gemma 4 E2B 5.1B Q4_K_M 2.9 GB 0.89 GB 4.4 GB
10.0 GB free
47 Ollama Run
Qwen3.5 4B 4.66B Q4_K_M 2.6 GB 4.00 GB 7.2 GB
7.2 GB free
52 Ollama Run
Gemma 3 4B 4.3B Q4_K_M 2.4 GB 3.11 GB 6.1 GB
8.3 GB free
56 Ollama Run
Qwen3 4B 4.02B Q4_K_M 2.3 GB 4.50 GB 7.4 GB
7.0 GB free
60 Ollama Run
LFM2.5 2.6B 2.7B Q4_K_M 1.5 GB 2.00 GB 4.1 GB
10.3 GB free
89 llama.cpp Run
Qwen3.5 2B 2.27B Q4_K_M 1.3 GB 1.50 GB 3.4 GB
11.0 GB free
106 Ollama Run
Qwen3 1.7B 1.72B Q4_K_M 1.0 GB 3.50 GB 5.1 GB
9.3 GB free
140 Ollama Run
Llama 3.2 1B Instruct 1.24B Q4_K_M 0.7 GB 4.00 GB 5.3 GB
9.1 GB free
194 Ollama Run
Gemma 3 1B 1.0B Q4_K_M 0.6 GB 0.14 GB 1.3 GB
13.1 GB free
241 Ollama Run
Qwen3.5 0.8B 0.87B Q4_K_M 0.5 GB 1.50 GB 2.6 GB
11.8 GB free
277 Ollama Run
Qwen3 0.6B 0.6B Q4_K_M 0.3 GB 3.50 GB 4.4 GB
10.0 GB free
402 Ollama Run

Runs — tight

2 models · fits at 128K, drop context or quant to go longer
ModelParamsQuantWeights +KV 128KTotal VRAM headroom Tok/sRuntime
Olmo 3 7B Instruct 7.3B Q4_K_M 4.1 GB 9.50 GB 14.2 GB
0.2 GB free
33 llama.cpp Run
Granite 4.1 3B 3.4B Q4_K_M 1.9 GB 10.00 GB 12.5 GB
1.9 GB free
71 llama.cpp Run

Won't fit in VRAM

57 models · some are possible with CPU offload
ModelParamsQuantWeights +KV 128KTotal Over budget Tok/sRuntime
Qwen3.5 397B-A17B 403B MoE / 17B act Q4_K_M 226.6 GB 3.75 GB 231.0 GB +216.6 GB ~1.5 llama.cpp Why
Ornith 1.5 397B-A17B 397B MoE / 17B act Q4_K_M 223.2 GB 3.75 GB 227.6 GB +213.2 GB ~1.5 llama.cpp Why
GLM-5.3-Flash 320B-A18B 321B MoE / 18B act Q4_K_M 180.5 GB 1.38 GB 182.5 GB +168.1 GB ~1.5 llama.cpp Why
DeepSeek V4 Flash 284B-A13B 284B MoE / 13B act Q4_K_M 159.7 GB 6.05 GB 166.3 GB +151.9 GB ~2.0 llama.cpp Why
Qwen3 235B-A22B 235B MoE / 22B act Q4_K_M 132.1 GB 23.50 GB 156.2 GB +141.8 GB ~1.1 llama.cpp Why
Qwen3.8-Flash-Next 180B-A6B 180B MoE / 6B act Q4_K_M 101.2 GB 3.00 GB 104.8 GB +90.4 GB ~4.5 llama.cpp Why
Qwen3.5 122B-A10B 125B MoE / 10B act Q4_K_M 70.3 GB 3.00 GB 73.9 GB +59.5 GB ~2.8 llama.cpp Why
Nemotron 3 Super 120B-A12B 124B MoE / 12B act Q4_K_M 69.7 GB 1.00 GB 71.3 GB +56.9 GB ~2.4 llama.cpp Why
Ling 3.0 Flash 124B-A5B 124B MoE / 5.1B act Q4_K_M 69.7 GB 0.98 GB 71.3 GB +56.9 GB ~5.6 llama.cpp Why
Mistral Small 4 119B-A6B 119B MoE / 6.5B act Q4_K_M 66.9 GB 2.81 GB 70.3 GB +55.9 GB ~4.3 llama.cpp Why
gpt-oss 120B 117B MoE / 5.1B act MXFP4 60.5 GB 4.50 GB 65.6 GB +51.2 GB ~6.0 llama.cpp Why
Llama 4 Scout 109B-A17B 109B MoE / 17B act Q4_K_M 61.3 GB 7.13 GB 69.0 GB +54.6 GB ~1.6 llama.cpp Why
Llama 3.3 70B Instruct 70.6B Q4_K_M 39.7 GB 40.00 GB 80.3 GB +65.9 GB ~0.9 llama.cpp Why
DeepSeek-R1-Distill-Llama 70B 70.6B Q4_K_M 39.7 GB 40.00 GB 80.3 GB +65.9 GB ~0.9 llama.cpp Why
Qwen3.6 35B-A3B 35.9B MoE / 3.3B act Q4_K_M 20.2 GB 2.50 GB 23.3 GB +8.9 GB ~13 llama.cpp Why
Ornith 1.5 35B-A3B 35.9B MoE / 3.3B act Q4_K_M 20.2 GB 2.50 GB 23.3 GB +8.9 GB ~13 llama.cpp Why
LLM-jp 4 33B Thinking 33.2B Q4_K_M 18.7 GB 16.00 GB 35.3 GB +20.9 GB ~1.9 llama.cpp Why
Qwen3 32B 32.8B Q4_K_M 18.4 GB 32.00 GB 51.0 GB +36.6 GB ~2.0 llama.cpp Why
DeepSeek-R1-Distill-Qwen 32B 32.8B Q4_K_M 18.4 GB 32.00 GB 51.0 GB +36.6 GB ~2.0 llama.cpp Why
Qwen2.5-Coder 32B 32.8B Q4_K_M 18.4 GB 32.00 GB 51.0 GB +36.6 GB ~2.0 llama.cpp Why
Olmo 3.1 32B Instruct 32.2B Q4_K_M 18.1 GB 4.75 GB 23.5 GB +9.1 GB ~3.2 llama.cpp Why
Nemotron 3.5 Lightning 30B-A3B 31.6B MoE / 3.2B act Q4_K_M 17.8 GB 0.75 GB 19.1 GB +4.7 GB ~17 llama.cpp Why
Gemma 4 31B 31.3B Q4_K_M 17.6 GB 20.78 GB 39.0 GB +24.6 GB ~2.1 llama.cpp Why
GLM-4.7-Flash 30B-A3B 31.2B MoE / 3B act Q4_K_M 17.5 GB 6.61 GB 24.8 GB +10.4 GB ~12 llama.cpp Why
Qwen3 30B-A3B 30.5B MoE / 3.3B act Q4_K_M 17.1 GB 12.00 GB 29.7 GB +15.3 GB ~8.2 llama.cpp Why
Qwen3 Coder 30B-A3B 30.5B MoE / 3.3B act Q4_K_M 17.1 GB 12.00 GB 29.7 GB +15.3 GB ~8.2 llama.cpp Why
Muse Glimmer 30B 29.8B Q4_K_M 16.8 GB 1.70 GB 19.1 GB +4.7 GB ~4.6 llama.cpp Why
Granite 4.1 30B 28.9B Q4_K_M 16.3 GB 32.00 GB 48.9 GB +34.5 GB ~2.2 llama.cpp Why
Qwen3.8 27B 27.8B Q4_K_M 15.6 GB 8.00 GB 24.2 GB +9.8 GB ~3.2 llama.cpp Why
Qwen3.6 27B 27.8B Q4_K_M 15.6 GB 8.00 GB 24.2 GB +9.8 GB ~3.2 llama.cpp Why
Gemma 3 27B 27.4B Q4_K_M 15.4 GB 10.41 GB 26.4 GB +12.0 GB ~2.8 llama.cpp Why
Gemma 4 26B-A4B 26.5B MoE / 3.8B act Q4_K_M 14.9 GB 5.20 GB 20.7 GB +6.3 GB ~11 llama.cpp Why
Devstral Small 2 24B 24B Q4_K_M 13.5 GB 20.00 GB 34.1 GB +19.7 GB ~2.7 llama.cpp Why
Mistral Small 3.2 24B 23.6B Q4_K_M 13.3 GB 20.00 GB 33.9 GB +19.5 GB ~2.7 llama.cpp Why
gpt-oss 20B 20.9B MoE / 3.6B act MXFP4 10.8 GB 3.00 GB 14.4 GB +0.0 GB ~28 llama.cpp Why
Qwen3 14B 14.8B Q4_K_M 8.3 GB 20.00 GB 28.9 GB +14.5 GB ~4.4 llama.cpp Why
DeepSeek-R1-Distill-Qwen 14B 14.8B Q4_K_M 8.3 GB 24.00 GB 32.9 GB +18.5 GB ~4.4 llama.cpp Why
Qwen2.5-Coder 14B 14.8B Q4_K_M 8.3 GB 24.00 GB 32.9 GB +18.5 GB ~4.4 llama.cpp Why
Ministral 3 14B 13.9B Q4_K_M 7.8 GB 20.00 GB 28.4 GB +14.0 GB ~4.6 llama.cpp Why
Gemma 3 12B 12.2B Q4_K_M 6.9 GB 8.31 GB 15.8 GB +1.4 GB ~13 llama.cpp Why
Mistral NeMo 12B 12.2B Q4_K_M 6.9 GB 20.00 GB 27.5 GB +13.1 GB ~5.3 llama.cpp Why
Gemma 4 12B 12B Q4_K_M 6.7 GB 8.31 GB 15.7 GB +1.3 GB ~13 llama.cpp Why
Ministral 3 8B 8.92B Q4_K_M 5.0 GB 17.00 GB 22.6 GB +8.2 GB ~7.2 llama.cpp Why
Granite 4.1 8B 8.79B Q4_K_M 4.9 GB 20.00 GB 25.5 GB +11.1 GB ~7.3 llama.cpp Why
Qwen3 8B 8.19B Q4_K_M 4.6 GB 18.00 GB 23.2 GB +8.8 GB ~7.9 llama.cpp Why
Llama 3.1 8B Instruct 8.03B Q4_K_M 4.5 GB 16.00 GB 21.1 GB +6.7 GB ~8.0 llama.cpp Why
Ministral 3 3B 3.85B Q4_K_M 2.2 GB 13.00 GB 15.8 GB +1.4 GB ~23 llama.cpp Why
Phi-4-mini 3.8B 3.84B Q4_K_M 2.2 GB 16.00 GB 18.8 GB +4.4 GB ~17 llama.cpp Why
Llama 3.2 3B Instruct 3.21B Q4_K_M 1.8 GB 14.00 GB 16.4 GB +2.0 GB ~20 llama.cpp Why
Kimi K3 2.8T-A104B 2780B MoE / 104B act Q4_K_M 1563.2 GB 3.38 GB 1567.1 GB +1552.7 GB ~0.2 llama.cpp Why
Qwen3.8 2.4T-A95B 2446B MoE / 95B act Q4_K_M 1375.4 GB 11.50 GB 1387.5 GB +1373.1 GB ~0.3 llama.cpp Why
DeepSeek V4 Pro 1.6T-A49B 1650B MoE / 49B act Q4_K_M 927.8 GB 8.58 GB 937.0 GB +922.6 GB ~0.5 llama.cpp Why
Kimi K2.6 1T-A32B 1027B MoE / 32B act Q4_K_M 577.5 GB 8.58 GB 586.6 GB +572.2 GB ~0.8 llama.cpp Why
GLM-5.2 744B-A40B 753B MoE / 40B act Q4_K_M 423.4 GB 10.97 GB 435.0 GB +420.6 GB ~0.6 llama.cpp Why
GLM-5.3 744B-A40B 753B MoE / 40B act Q4_K_M 423.4 GB 10.97 GB 435.0 GB +420.6 GB ~0.6 llama.cpp Why
DeepSeek-R1 671B 671B MoE / 37B act Q4_K_M 377.3 GB 8.58 GB 386.5 GB +372.1 GB ~0.7 llama.cpp Why
DeepSeek V4.1 Flash 552B-A16B New 552B MoE / 16B act Q4_K_M 310.4 GB 0.57 GB 311.5 GB +297.1 GB ~1.6 vLLM Why

Totals = quantised weights + f16 KV cache at 128K tokens + 0.6 GB runtime overhead, against 14.4 GB usable VRAM (16 GB less display and driver reserve). Tokens per second are estimated from 224 GB/s peak memory bandwidth at batch 1 — not measured. Read the method.


Same memory, different speed

10 of the 19 devices have the same 14.4 GB usable, so they run exactly the models above. Generation speed scales with memory bandwidth.

DeviceBandwidthSpeed vs yours
GeForce RTX 5080
Blackwell · 2025
960 GB/s 4.29×
GeForce RTX 5070 Ti
Blackwell · 2025
896 GB/s 4.00×
CMP 100-210
Volta · 2021
829 GB/s 3.70×
GeForce RTX 4080 Super
Ada Lovelace · 2024
736 GB/s 3.29×
Tesla P100 16 GB
Pascal · 2016
732 GB/s 3.27×
GeForce RTX 4080
Ada Lovelace · 2022
717 GB/s 3.20×
GeForce RTX 4070 Ti Super
Ada Lovelace · 2024
672 GB/s 3.00×
Radeon RX 9070 XT
RDNA 4 · 2025
645 GB/s 2.88×
Radeon RX 9070
RDNA 4 · 2025
645 GB/s 2.88×
Radeon RX 7800 XT
RDNA 3 · 2023
624 GB/s 2.79×

Best LLM for 16 GB VRAM, with every card in the class →

What more memory unlocks

At Q4_K_M and 8K context this device fits 38 of 80 models. Each step up, and the largest models it adds:

Every device ranked by what it can run →