FitMyLLM
NVIDIA· KEPLER 2.0

NVIDIA Jetson TK1

Running LLMs on the Jetson TK1 — the long read: which models fit at which quantisation, and the settings worth changing. · Or what a budget buys

VRAM
2 GB
BUDGET
BANDWIDTH
7
GB/S
MODELS Q4
47/449
10%
7B Q4 SPEED
~1
SLOW
▸ MODEL COVERAGE @ Q410% OF ALL
▸ ESTIMATED SPEED· BY MODEL SIZE @ Q4

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

3B
1.7GB NEEDED
7B
3.9GB NEEDED
14B
7.9GB NEEDED
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
2 GB
BANDWIDTH
7 GB/s
FP16 COMPUTE
0.4 TFLOPS
TDP
8W
MEMORY
DDR3L
ARCHITECTURE
Kepler 2.0
CUDA CORES
192
PCIE
Gen 1 x1
4
FAST MODELS · >30 TOK/S
Real-time chat speed
18
USABLE · >10 TOK/S
Comfortable for all tasks
47
TOTAL COMPATIBLE
Fit in VRAM at Q4
▸ DON’T WANT TO BUY?

Test Jetson TK1 (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· 47
S
Falcon-H1R Tiny 90M0.09B
FALCON·256K CTX· CHAT· REASONING
62
TOK/S · 27% VRAM
A
SmolLM2 135M0.135B
SMOLLM·2K CTX· CHAT
41
TOK/S · 29% VRAM
A
nomic-embed-text-v1.5 100M0.14B
EMBEDDING·8K CTX· CHAT
40
TOK/S · 29% VRAM
A
GPT-2 124M0.14B
GPT2·1K CTX· CHAT
40
TOK/S · 29% VRAM
C
SmolVLM 256M0.256B
SMOLLM·8K CTX· CHAT· VISION
22
TOK/S · 32% VRAM
C
Gemma 3 270M0.27B
GEMMA·32K CTX· CHAT
21
TOK/S · 33% VRAM
C
Snowflake Arctic Embed M v2.00.305B
EMBEDDING·8K CTX· EMBEDDING· MULTILINGUAL
18
TOK/S · 34% VRAM
C
bge-large-en-v1.5 335M0.335B
EMBEDDING·1K CTX· CHAT
17
TOK/S · 35% VRAM
C
mxbai-embed-large-v10.335B
EMBEDDING·1K CTX· EMBEDDING
17
TOK/S · 35% VRAM
C
Snowflake Arctic Embed L0.335B
EMBEDDING·1K CTX· EMBEDDING
17
TOK/S · 35% VRAM
C
SmolLM2 360M0.36B
SMOLLM·8K CTX· CHAT
16
TOK/S · 35% VRAM
C
LFM2 350M0.35B
LFM·125K CTX· CHAT· TOOL_USE
16
TOK/S · 35% VRAM
C
GPT-2 Medium 345M0.38B
GPT2·1K CTX· CHAT
15
TOK/S · 36% VRAM
D
Falcon-H1 0.5B0.52B
FALCON·128K CTX· CHAT
11
TOK/S · 40% VRAM
D
Qwen 2.5 0.5B0.5B
QWEN·32K CTX· CHAT
11
TOK/S · 40% VRAM
D
SmolVLM 500M0.5B
SMOLLM·8K CTX· CHAT· VISION
11
TOK/S · 40% VRAM
D
BGE-M30.568B
EMBEDDING·8K CTX· EMBEDDING
10
TOK/S · 42% VRAM
D
Snowflake Arctic Embed L v2.00.568B
EMBEDDING·8K CTX· EMBEDDING· MULTILINGUAL
10
TOK/S · 42% VRAM
D
Qwen 1.5 0.5B0.62B
QWEN·32K CTX· CHAT
9
TOK/S · 43% VRAM
D
Qwen3-Embedding 0.6B0.6B
EMBEDDING·32K CTX· EMBEDDING· MULTILINGUAL
9
TOK/S · 43% VRAM
D
Falcon-H1R Tiny 0.6B0.6B
FALCON·32K CTX· CHAT· REASONING
9
TOK/S · 43% VRAM
D
Falcon Perception 0.6B0.6B
FALCON·4K CTX· VISION
9
TOK/S · 43% VRAM
D
LFM2 700M0.74B
LFM·125K CTX· CHAT· TOOL_USE
8
TOK/S · 47% VRAM
F
GPT-2 Large 774M0.81B
GPT2·1K CTX· CHAT
7
TOK/S · 49% VRAM
F
Qwen3 0.6B0.75B
QWEN·32K CTX· CHAT· REASONING
7
TOK/S · 47% VRAM
F
InternVL3 1B0.94B
OTHER·32K CTX· CHAT· VISION
6
TOK/S · 53% VRAM
F
Qwen 3.5 0.8B0.87B
QWEN·256K CTX· CHAT· CODING· MULTILINGUAL
6
TOK/S · 51% VRAM
F
LFM2.5-1.2B-Thinking1.2B
LFM·122K CTX· CHAT· REASONING· TOOL_USE
5
TOK/S · 61% VRAM
F
Llama-3.2-1B1.2B
LLAMA·4K CTX· CHAT
5
TOK/S · 61% VRAM
F
LFM2 1.2B1.2B
LFM·125K CTX· CHAT· TOOL_USE· MULTILINGUAL
5
TOK/S · 61% VRAM
F
Zamba2 1.2B1.2B
OTHER·4K CTX· CHAT
5
TOK/S · 61% VRAM
F
TinyLlama 1.1B1.1B
LLAMA·2K CTX· CHAT
5
TOK/S · 58% VRAM
F
MiniCPM5 1B1.08B
MINICPM·128K CTX· CHAT· REASONING· MULTILINGUAL
5
TOK/S · 57% VRAM
F
stablelm-2-1_6b1.6B
STABLELM·4K CTX· CHAT
4
TOK/S · 73% VRAM
F
Falcon-H1 1.5B1.55B
FALCON·128K CTX· CHAT· CODING
4
TOK/S · 72% VRAM
F
Qwen2.5-Coder-1.5B1.5B
QWEN·32K CTX· CHAT· TOOL_USE· CODING
4
TOK/S · 70% VRAM
F
Qwen2 Math 1.5B1.5B
QWEN·4K CTX· REASONING
4
TOK/S · 70% VRAM
F
Qwen 2.5 1.5B1.5B
QWEN·32K CTX· CHAT· CODING
4
TOK/S · 70% VRAM
F
Yi Coder 1.5B1.5B
YI·125K CTX· CODING
4
TOK/S · 70% VRAM
F
Stella en 1.5B v51.5B
EMBEDDING·128K CTX· EMBEDDING
4
TOK/S · 70% VRAM
▸ NEXT STEP

Get personalized recommendations.

See ranked models with benchmark scores, run commands, and precise speed estimates for your Jetson TK1.

WHAT THIS CARD IS WORTH

Jetson TK1 holds 47 of the models in our catalogue and is, in practice, a Q4_K_M card — the largest it takes is Moondream2 1.9B at Q4_K_M.

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
Moondream2 1.9B1.9B · Q4_K_M1.7 GB5 tok/sEST
Falcon3-1B1.67B · Q5_K_M1.8 GB4 tok/sEST
LFM2-VL 1.6B1.6B · Q5_K_M1.7 GB5 tok/sEST
Falcon-H1 1.5B1.55B · Q6_K1.7 GB4 tok/sEST
Qwen2.5-Coder-1.5B1.5B · Q6_K1.6 GB4 tok/sEST
Qwen2 Math 1.5B1.5B · Q6_K1.6 GB4 tok/sEST
Qwen 2.5 1.5B1.5B · Q6_K1.6 GB4 tok/sEST
Stella en 1.5B v51.5B · Q6_K1.6 GB4 tok/sEST
AGAINST CARDS OF SIMILAR MEMORY
M2 (8GB)5 GB$5992.8 tok/s per $100
M3 (8GB)5 GB$5992.8 tok/s per $100
M1 (8GB)5 GB$4992.2 tok/s per $100
RTX 2060 6GB6 GB$15026.7 tok/s per $100
Jetson TK12 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 Jetson TK1 — 2 GB VRAM.

JETSON TK1 SPEC
BRAND
NVIDIA
VRAM
2 GB DDR3L
BANDWIDTH
7 GB/s
FP16 COMPUTE
0.4 TFLOPS
FP32 COMPUTE
0.4 TFLOPS
CUDA CORES
192
TDP
8 W
ARCHITECTURE
Kepler 2.0
▸ AI CAPABILITY
47/ 449 models @ Q4

With 2 GB VRAM and 7 GB/s bandwidth, this GPU handles models up to 1.61B parameters.

Speed ≈ bandwidth / model_size × efficiency. A 7B model at Q4 runs at ~1 tok/s.

§ 01TOP MODELS FOR JETSON TK1
47 FIT · SHOWING 20
MODELSIZEVRAM Q4TOK/SAVG
GPT-2 XL 1.5B1.61B1.5 GB35.1
stablelm-2-1_6b1.6B1.5 GB49.5
Falcon-H1 1.5B1.55B1.4 GB443.8
Qwen2.5-Coder-1.5B1.5B1.4 GB419.6
Qwen2 Math 1.5B1.5B1.4 GB419.6
Qwen 2.5 1.5B1.5B1.4 GB430.2
Yi Coder 1.5B1.5B1.4 GB414.6
Stella en 1.5B v51.5B1.4 GB4
Phi-1 1.3B1.42B1.4 GB47.2
Phi-1.5 1.3B1.42B1.4 GB47.2
DeepSeek Coder 1.3B1.35B1.3 GB416.8
EXAONE-4.0-1.2B1.3B1.3 GB418.9
OPT 1.3B1.3B1.3 GB45.3
MiniCPM-V 4.61.3B1.3 GB421.5
LFM2.5-1.2B-Thinking1.2B1.2 GB519.6
Llama-3.2-1B1.2B1.2 GB510.1
LFM2 1.2B1.2B1.2 GB515.6
Zamba2 1.2B1.2B1.2 GB541.5
TinyLlama 1.1B1.1B1.2 GB513.6
MiniCPM5 1B1.08B1.1 GB526.5