FitMyLLM
NVIDIA· ADA LOVELACE

NVIDIA RTX 5000 Embedded Ada Generation X2

Running LLMs on the RTX 5000 Embedded Ada Generation X2 — the long read: which models fit at which quantisation, and the settings worth changing. · Or what a budget buys

VRAM
16 GB
MID-RANGE
BANDWIDTH
576
GB/S
MODELS Q4
262/449
58%
7B Q4 SPEED
~66
BLAZING
▸ MODEL COVERAGE @ Q458% OF ALL
▸ ESTIMATED SPEED· BY MODEL SIZE @ Q4

Average speeds at Q4 quantization. Actual performance varies by model architecture and context length.

3B
~154
TOK/S
7B
~66
TOK/S
14B
~33
TOK/S
32B
18.0GB NEEDED
70B
39.4GB NEEDED
▸ MEASURED RIG REPORTS

We rent the machine and time every model on it: decode, VRAM peak, concurrency, watts and cost per million tokens. Yours may already be one of them — and two are free to read in full.

SEE THE REPORTS
▸ SPECIFICATIONS
VRAM
16 GB
BANDWIDTH
576 GB/s
FP16 COMPUTE
32.7 TFLOPS
TDP
150W
MEMORY
GDDR6
ARCHITECTURE
Ada Lovelace
CUDA CORES
9,728
TENSOR CORES
304
PCIE
Gen 4 x16
259
FAST MODELS · >30 TOK/S
Real-time chat speed
262
USABLE · >10 TOK/S
Comfortable for all tasks
262
TOTAL COMPATIBLE
Fit in VRAM at Q4
▸ DON’T WANT TO BUY?

Test RTX 5000 Embedded Ada Generation X2 (or anything bigger) without committing. Pay by the second, cancel anytime.

Spin up in ~60s. Pay by the second. Cancel anytime.

Some links are affiliate links — we may earn a small commission at no extra cost to you. This helps keep FitMyLLM free and independent.

▸ COMPATIBLE MODELS· 262
S
Falcon-H1R Tiny 90M0.09B
FALCON·256K CTX· CHAT· REASONING
5120
TOK/S · 3% VRAM
S
SmolLM2 135M0.135B
SMOLLM·2K CTX· CHAT
3413
TOK/S · 4% VRAM
S
nomic-embed-text-v1.5 100M0.14B
EMBEDDING·8K CTX· CHAT
3291
TOK/S · 4% VRAM
S
GPT-2 124M0.14B
GPT2·1K CTX· CHAT
3291
TOK/S · 4% VRAM
S
SmolVLM 256M0.256B
SMOLLM·8K CTX· CHAT· VISION
1800
TOK/S · 4% VRAM
S
Gemma 3 270M0.27B
GEMMA·32K CTX· CHAT
1707
TOK/S · 4% VRAM
S
Snowflake Arctic Embed M v2.00.305B
EMBEDDING·8K CTX· EMBEDDING· MULTILINGUAL
1511
TOK/S · 4% VRAM
S
bge-large-en-v1.5 335M0.335B
EMBEDDING·1K CTX· CHAT
1376
TOK/S · 4% VRAM
S
mxbai-embed-large-v10.335B
EMBEDDING·1K CTX· EMBEDDING
1376
TOK/S · 4% VRAM
S
Snowflake Arctic Embed L0.335B
EMBEDDING·1K CTX· EMBEDDING
1376
TOK/S · 4% VRAM
S
LFM2 350M0.35B
LFM·125K CTX· CHAT· TOOL_USE
1317
TOK/S · 4% VRAM
S
SmolLM2 360M0.36B
SMOLLM·8K CTX· CHAT
1280
TOK/S · 4% VRAM
S
GPT-2 Medium 345M0.38B
GPT2·1K CTX· CHAT
1213
TOK/S · 5% VRAM
S
Qwen 2.5 0.5B0.5B
QWEN·32K CTX· CHAT
922
TOK/S · 5% VRAM
S
SmolVLM 500M0.5B
SMOLLM·8K CTX· CHAT· VISION
922
TOK/S · 5% VRAM
S
Falcon-H1 0.5B0.52B
FALCON·128K CTX· CHAT
886
TOK/S · 5% VRAM
S
BGE-M30.568B
EMBEDDING·8K CTX· EMBEDDING
811
TOK/S · 5% VRAM
S
Snowflake Arctic Embed L v2.00.568B
EMBEDDING·8K CTX· EMBEDDING· MULTILINGUAL
811
TOK/S · 5% VRAM
S
Qwen3-Embedding 0.6B0.6B
EMBEDDING·32K CTX· EMBEDDING· MULTILINGUAL
768
TOK/S · 5% VRAM
S
Falcon-H1R Tiny 0.6B0.6B
FALCON·32K CTX· CHAT· REASONING
768
TOK/S · 5% VRAM
S
Falcon Perception 0.6B0.6B
FALCON·4K CTX· VISION
768
TOK/S · 5% VRAM
S
Qwen 1.5 0.5B0.62B
QWEN·32K CTX· CHAT
743
TOK/S · 5% VRAM
S
LFM2 700M0.74B
LFM·125K CTX· CHAT· TOOL_USE
623
TOK/S · 6% VRAM
S
Qwen3 0.6B0.75B
QWEN·32K CTX· CHAT· REASONING
614
TOK/S · 6% VRAM
S
GPT-2 Large 774M0.81B
GPT2·1K CTX· CHAT
569
TOK/S · 6% VRAM
S
Qwen 3.5 0.8B0.87B
QWEN·256K CTX· CHAT· CODING· MULTILINGUAL
530
TOK/S · 6% VRAM
S
InternVL3 1B0.94B
OTHER·32K CTX· CHAT· VISION
490
TOK/S · 7% VRAM
S
MiniCPM5 1B1.08B
MINICPM·128K CTX· CHAT· REASONING· MULTILINGUAL
427
TOK/S · 7% VRAM
S
TinyLlama 1.1B1.1B
LLAMA·2K CTX· CHAT
419
TOK/S · 7% VRAM
S
LFM2.5-1.2B-Thinking1.2B
LFM·122K CTX· CHAT· REASONING· TOOL_USE
384
TOK/S · 8% VRAM
S
Llama-3.2-1B1.2B
LLAMA·4K CTX· CHAT
384
TOK/S · 8% VRAM
S
LFM2 1.2B1.2B
LFM·125K CTX· CHAT· TOOL_USE· MULTILINGUAL
384
TOK/S · 8% VRAM
S
Zamba2 1.2B1.2B
OTHER·4K CTX· CHAT
384
TOK/S · 8% VRAM
S
EXAONE-4.0-1.2B1.3B
EXAONE·64K CTX· CHAT
354
TOK/S · 8% VRAM
S
OPT 1.3B1.3B
OPT·2K CTX· CHAT
354
TOK/S · 8% VRAM
S
MiniCPM-V 4.61.3B
OTHER·256K CTX· CHAT· VISION
354
TOK/S · 8% VRAM
S
DeepSeek Coder 1.3B1.35B
DEEPSEEK·16K CTX· CODING
341
TOK/S · 8% VRAM
S
Phi-1 1.3B1.42B
PHI·2K CTX· CODING
325
TOK/S · 8% VRAM
S
Phi-1.5 1.3B1.42B
PHI·2K CTX· CHAT· CODING
325
TOK/S · 8% VRAM
S
LFM2 8B A1B8.3BMoE
LFM·125K CTX· CHAT· TOOL_USE· MULTILINGUAL
307
TOK/S · 35% VRAM
▸ NEXT STEP

Get personalized recommendations.

See ranked models with benchmark scores, run commands, and precise speed estimates for your RTX 5000 Embedded Ada Generation X2.

WHAT THIS CARD IS WORTH

RTX 5000 Embedded Ada Generation X2 holds 262 of the models in our catalogue and is, in practice, a IQ4_XS card — the largest it takes is ERNIE 4.5 21B A3B at IQ4_XS.

TOKENS/SEC PER $100
8B at Q4_K_M, so cards compare like for like
VRAM PER $100
what memory costs on this card
THE BIGGEST IT TAKES
ERNIE 4.5 21B A3B21.95B · IQ4_XS13.9 GB163 tok/sEST
GPT-OSS 20B21B · Q4_K_S13.8 GB130 tok/sEST
Reka Flash 321B · Q4_K_S14.2 GB22 tok/sEST
Reka Flash 3.121B · Q4_K_S14.2 GB22 tok/sEST
InternLM2 20B19.8B · Q4_K_M14.3 GB23 tok/sEST
InternLM2.5 20B19.8B · Q4_K_M14.3 GB23 tok/sEST
Ling-lite 16.8B16.8B · Q5_K_M13.6 GB160 tok/sEST
DeepSeek V2 Lite 16B16B · Q5_K_M12.9 GB160 tok/sEST
AGAINST CARDS OF SIMILAR MEMORY
M1 Max (32GB)21 GB$1,4993.0 tok/s per $100
M2 Max (32GB)21 GB$1,7992.5 tok/s per $100
M2 Pro (32GB)21 GB$1,4991.9 tok/s per $100
M4 (32GB)21 GB$1,1991.7 tok/s per $100
RTX 5000 Embedded Ada Generation X216 GB

Neighbours in memory rather than in price: memory decides whether a card can do the job at all, so two cards of the same size at different prices is the comparison you are making. Ordered by memory, then bandwidth — not by the value column, which is worked out from bandwidth and price alone and therefore rewards a cheap card whatever its software stack does to that bandwidth in practice. Read it as one input, not as a ranking.

▸ DEVICE UNDER TEST

NVIDIA RTX 5000 Embedded Ada Generation X2 — 16 GB VRAM.

RTX 5000 EMBEDDED ADA GENERATION X2 SPEC
BRAND
NVIDIA
VRAM
16 GB GDDR6
BANDWIDTH
576 GB/s
FP16 COMPUTE
32.7 TFLOPS
FP32 COMPUTE
32.7 TFLOPS
CUDA CORES
9,728
TENSOR CORES
304
TDP
150 W
ARCHITECTURE
Ada Lovelace
▸ AI CAPABILITY
262/ 449 models @ Q4

With 16 GB VRAM and 576 GB/s bandwidth, this GPU handles models up to 19.8B parameters.

Speed ≈ bandwidth / model_size × efficiency. A 7B model at Q4 runs at ~66 tok/s.

RENT IT — LIVE PRICES
Checking what the clouds are charging…
§ 01TOP MODELS FOR RTX 5000 EMBEDDED ADA GENERATION X2
262 FIT · SHOWING 20
MODELSIZEVRAM Q4TOK/SAVG
InternLM2 20B19.8B12.6 GB2345.1
InternLM2.5 20B19.8B12.6 GB2350.9
Ling-lite 16.8B16.8B10.8 GB192
DeepSeek V2 Lite 16B16B10.3 GB19238.0
StarCoder2 15B15.96B10.2 GB2926.5
DeepSeek-Coder-V2-Lite 15.7B15.7B10.1 GB19243.0
DeepSeek-VL2 Small 16B15.7B10.1 GB19243.1
StarCoder 15B15.5B10.0 GB3021.0
InternVL3 14B15.12B9.7 GB3038.1
Phi-4-reasoning-vision 15B15B9.7 GB3142.8
DeepSeek R1 Distill Qwen 14B14.8B9.5 GB3143.9
DeepCoder 14B14.8B9.5 GB3138.7
Qwen2.5-Coder-14B14.8B9.5 GB3141.3
Qwen2.5-14B14.8B9.5 GB3141.3
Qwen3 14B14.8B9.5 GB3145.7
phi-4 14B14.66B9.4 GB3133.7
Phi-4-reasoning 14B14.66B9.4 GB3133.7
Phi-4-reasoning-plus 14B14.66B9.4 GB3175.5
Ministral 3 14B14B9.0 GB3325.9
Phi-3-medium-14b14B9.0 GB3333.7