FitMyLLM
NVIDIA· Volta

NVIDIA TITAN V

Running LLMs on the TITAN V — the long read: which models fit at which quantisation, and the settings worth changing. · Or what a budget buys

VRAM
12 GB
Entry-level
Bandwidth
651
GB/s
Models Q4
263/498
53%
7B Q4 speed
~74
Blazing
Model coverage at Q453% of all
Estimated speed· By model size at Q4

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

3B
~174
tok/s
7B
~74
tok/s
14B
~37
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
12 GB
Bandwidth
651 GB/s
FP16 Compute
29.8 TFLOPS
TDP
250W
Memory
HBM2
Architecture
Volta
CUDA Cores
5,120
Tensor Cores
640
PCIe
Gen 3 x16
263
Fast models · >30 tok/s
Real-time chat speed
263
Usable · >10 tok/s
Comfortable for all tasks
263
Total compatible
Fit in VRAM at Q4
Don’t want to buy?

Test TITAN V (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· 263
S
Falcon-H1R Tiny 90M0.09B
Falcon·256K ctx· chat· reasoning
5787
tok/s · 5% VRAM
›
S
SmolLM2 135M0.135B
Smollm·2K ctx· chat
3858
tok/s · 5% VRAM
›
S
nomic-embed-text-v1.5 100M0.14B
Embedding·8K ctx· chat
3720
tok/s · 5% VRAM
›
S
GPT-2 124M0.14B
Gpt2·1K ctx· chat
3720
tok/s · 5% VRAM
›
S
SmolVLM 256M0.256B
Smollm·8K ctx· chat· vision
2034
tok/s · 5% VRAM
›
S
Gemma 3 270M0.27B
Gemma·32K ctx· chat
1929
tok/s · 5% VRAM
›
S
Snowflake Arctic Embed M v2.00.305B
Embedding·8K ctx· embedding· multilingual
1708
tok/s · 6% VRAM
›
S
bge-large-en-v1.5 335M0.335B
Embedding·1K ctx· chat
1555
tok/s · 6% VRAM
›
S
mxbai-embed-large-v10.335B
Embedding·1K ctx· embedding
1555
tok/s · 6% VRAM
›
S
Snowflake Arctic Embed L0.335B
Embedding·1K ctx· embedding
1555
tok/s · 6% VRAM
›
S
LFM2 350M0.35B
Lfm·125K ctx· chat· tool_use
1488
tok/s · 6% VRAM
›
S
SmolLM2 360M0.36B
Smollm·8K ctx· chat
1447
tok/s · 6% VRAM
›
S
GPT-2 Medium 345M0.38B
Gpt2·1K ctx· chat
1371
tok/s · 6% VRAM
›
S
Qwen 2.5 0.5B0.5B
Qwen·32K ctx· chat
1042
tok/s · 7% VRAM
›
S
SmolVLM 500M0.5B
Smollm·8K ctx· chat· vision
1042
tok/s · 7% VRAM
›
S
Falcon-H1 0.5B0.52B
Falcon·128K ctx· chat
1002
tok/s · 7% VRAM
›
S
BGE-M30.568B
Embedding·8K ctx· embedding
917
tok/s · 7% VRAM
›
S
Snowflake Arctic Embed L v2.00.568B
Embedding·8K ctx· embedding· multilingual
917
tok/s · 7% VRAM
›
S
Qwen3-Embedding 0.6B0.6B
Embedding·32K ctx· embedding· multilingual
868
tok/s · 7% VRAM
›
S
Falcon-H1R Tiny 0.6B0.6B
Falcon·32K ctx· chat· reasoning
868
tok/s · 7% VRAM
›
S
Falcon Perception 0.6B0.6B
Falcon·4K ctx· vision
868
tok/s · 7% VRAM
›
S
Qwen 1.5 0.5B0.62B
Qwen·32K ctx· chat
840
tok/s · 7% VRAM
›
S
LFM2 700M0.74B
Lfm·125K ctx· chat· tool_use
704
tok/s · 8% VRAM
›
S
Qwen3 0.6B0.75B
Qwen·32K ctx· chat· reasoning
694
tok/s · 8% VRAM
›
S
GPT-2 Large 774M0.81B
Gpt2·1K ctx· chat
643
tok/s · 8% VRAM
›
S
Qwen 3.5 0.8B0.87B
Qwen·256K ctx· chat· coding· multilingual
599
tok/s · 9% VRAM
›
S
InternVL3 1B0.94B
Other·32K ctx· chat· vision
554
tok/s · 9% VRAM
›
S
MiniCPM5 1B1.08B
Minicpm·128K ctx· chat· reasoning· multilingual
482
tok/s · 10% VRAM
›
S
K2 Horizon 0.9B1.08B
Other·128K ctx· chat· coding· reasoning
482
tok/s · 10% VRAM
›
S
TinyLlama 1.1B1.1B
Llama·2K ctx· chat
473
tok/s · 10% VRAM
›
S
LFM2.5-1.2B-Thinking1.2B
Lfm·122K ctx· chat· reasoning· tool_use
434
tok/s · 10% VRAM
›
S
Llama-3.2-1B1.2B
Llama·4K ctx· chat
434
tok/s · 10% VRAM
›
S
LFM2 1.2B1.2B
Lfm·125K ctx· chat· tool_use· multilingual
434
tok/s · 10% VRAM
›
S
Zamba2 1.2B1.2B
Other·4K ctx· chat
434
tok/s · 10% VRAM
›
S
Ling 3.0 Tiny7.89BMoE
Other·128K ctx· chat· coding· reasoning
401
tok/s · 44% VRAM
›
S
EXAONE-4.0-1.2B1.3B
Exaone·64K ctx· chat
401
tok/s · 11% VRAM
›
S
OPT 1.3B1.3B
Opt·2K ctx· chat
401
tok/s · 11% VRAM
›
S
MiniCPM-V 4.61.3B
Other·256K ctx· chat· vision
401
tok/s · 11% VRAM
›
S
DeepSeek Coder 1.3B1.35B
Deepseek·16K ctx· coding
386
tok/s · 11% VRAM
›
S
Phi-1 1.3B1.42B
Phi·2K ctx· coding
367
tok/s · 11% VRAM
›
Next step

Get personalized recommendations.

See ranked models with benchmark scores, run commands, and precise speed estimates for your TITAN V.

What this card is worth

TITAN V holds 263 of the models in our catalogue and is, in practice, a IQ4_XS card — the largest it takes is Ling-lite 16.8B 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
Ling-lite 16.8B16.8B · IQ4_XS10.7 GB225 tok/sest
DeepSeek V2 Lite 16B16B · Q4_K_S10.6 GB215 tok/sest
StarCoder2 15B15.96B · Q4_K_S10.7 GB32 tok/sest
DeepSeek-Coder-V2-Lite 15.7B15.7B · Q4_K_S10.4 GB215 tok/sest
DeepSeek R1 Distill Qwen 14B14.8B · Q4_K_S10.4 GB35 tok/sest
DeepCoder 14B14.8B · Q4_K_S10.4 GB35 tok/sest
Qwen2.5-Coder-14B14.8B · Q4_K_S10.4 GB35 tok/sest
Qwen2.5-14B14.8B · Q4_K_S10.4 GB35 tok/sest
Against cards of similar memory
M4 Pro (24GB)16 GB$1,3992.5 tok/s per $100
M4 (24GB)16 GB$6992.9 tok/s per $100
M2 (24GB)16 GB$9991.7 tok/s per $100
M3 (24GB)16 GB$9991.7 tok/s per $100
TITAN V12 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 TITAN V — 12 GB VRAM.

TITAN V spec
Brand
NVIDIA
VRAM
12 GB HBM2
Bandwidth
651 GB/s
FP16 Compute
29.8 TFLOPS
FP32 Compute
14.9 TFLOPS
CUDA Cores
5,120
Tensor Cores
640
TDP
250 W
Architecture
Volta
AI capability
263/ 498 models @ Q4

With 12 GB VRAM and 651 GB/s bandwidth, this GPU handles models up to 14.8B parameters.

Speed ≈ bandwidth / model_size × efficiency. A 7B model at Q4 runs at ~74 tok/s.

Rent it — live prices
Checking what the clouds are charging…
Top models for TITAN V
263 fit · showing 20
ModelSizeVRAM Q4tok/sAvg
DeepSeek R1 Distill Qwen 14B14.8B9.5 GB3543.9
DeepCoder 14B14.8B9.5 GB3536.0
Qwen2.5-Coder-14B14.8B9.5 GB3541.3
Qwen2.5-14B14.8B9.5 GB3541.3
Qwen3 14B14.8B9.5 GB3545.7
phi-4 14B14.66B9.4 GB3633.7
Phi-4-reasoning 14B14.66B9.4 GB3633.7
Phi-4-reasoning-plus 14B14.66B9.4 GB3672.2
Phi-3-medium-14b14B9.0 GB3733.7
Qwen 1.5 14B14B9.0 GB3741.3
Ministral 3 14B Reasoning13.95B9.0 GB3773.6
Baichuan2 13B13B8.4 GB4023.6
Llama 2 13B13B8.4 GB4017.2
CodeLlama 13B13B8.4 GB4019.7
Vicuna 13B13B8.4 GB4011.8
LLaMA 1 13B13B8.4 GB4032.9
OPT 13B13B8.4 GB4035.8
Orca 2 13B13B8.4 GB4025.4
WizardCoder Python 13B13B8.4 GB4060.1
WizardLM 13B13B8.4 GB4019.5