llmrun

NVIDIA · Turing

Best AI Models for NVIDIA GeForce RTX 2080 Ti (11 GB)

Runs 7–9B models well; 14B only just fits.

11 GB GDDR6616 GB/s57 TFLOPS FP16250 WTuring 2018
Buyabout $999

Turing 11 GB: has Tensor Cores and flash-attention (cc 7.5) — the cheapest used card that still runs modern FP16 kernels well.

What LLMs can the RTX 2080 Ti run?

#ModelSpeedllmrun ScoreQuant · VRAM
1
Qwen3.5 9BAlibabaHow to run →
Q4_K_M · 6.6 GB · 63.0 t/s
63.0 tok/sQ4_K_M · 6.6 GB
2
Qwen3.5 4BAlibaba
Q4_K_M · 3.4 GB · 123 t/s
123 tok/sQ4_K_M · 3.4 GB
3
Gemma 2 9B ITGoogle
Q4_K_M · 7.3 GB · 65.6 t/s
65.6 tok/sQ4_K_M · 7.3 GB
4
Gemma 3 4B ITGoogle
Q4_K_M · 5.0 GB · 141 t/s
141 tok/sQ4_K_M · 5.0 GB
5
Ling 3.0 TinyInclusion AI
Q4_K_M · 5.5 GB · 274 t/s
274 tok/sQ4_K_M · 5.5 GB
See all 265 models on the RTX 2080 Ti →Best quant that fits · speeds estimated on this card

About the RTX 2080 Ti

The RTX 2080 Ti is one of the fastest 10–11 GB cards: it runs 7–9B models well. It is the most efficient per watt in its tier, and at its $999 MSRP it cost more per token than most of the cards compared. Turing generation, released 2018.

Speed in 10–11 GB#2616 GB/s
Per watt#10.25 tok/s per W on Qwen3.5 9B (9B params)
Per dollar#36.3 tok/s per $100 on Qwen3.5 9B (9B params)
GenerationTuringreleased 2018

RTX 2080 Ti strengths and weaknesses

Strengths

  • Runs 14B models: Qwen3 14B at 46.7 tok/s in 10.9 GB

Weaknesses

  • 27–31B models don't fit: Qwen3.8 27B needs 17.9 GB at Q4 vs 11 GB here
  • Little room for context: Qwen3 14B uses 10.9 of 11 GB, 0.1 GB left
  • No FP8 or NVFP4: Turing predates both; use GGUF, AWQ or GPTQ builds

Speed per dollar and per watt

On Qwen3.5 9B (9B params) at Q4_K_M. New prices are compared with new, used with used. The last two columns are the speed you get per $100 of price and per 100 W of board power; higher is better. The RTX 2080 Ti has no current price, so it is shown at its MSRP.

CardNew pricetok/stok/s per $100tok/s per 100 W
RTX 50608 GBBest new value
about $46045.8
10.0
31.6
RTX 507012 GB
about $84068.7
8.2
27.5
RTX 2080 Ti11 GB
about $99963.0
6.3
25.2
RTX 40608 GB
about $52027.8
5.3
24.2
CardUsed pricetok/stok/s per $100tok/s per 100 W
RTX 50608 GBBest used value
about $38845.8
11.8
31.6
RTX 3060 12GB12 GB
about $32936.8
11.2
21.6
RTX 507012 GB
about $73668.7
9.3
27.5
RTX 407012 GB
about $57551.5
9.0
25.8
RTX 4060 Ti 8GB8 GB
about $36229.4
8.1
18.4
RTX 4070 SUPER12 GB
about $67551.5
7.6
23.4
RTX 4070 Ti12 GB
about $72251.5
7.1
18.1
RTX 309024 GB
about $1,50095.7
6.4
27.3

Cost per token: the RTX 2080 Ti against 8–24 GB cards

On Qwen3.5 9B (9B params, Q4_K_M, 4K context). Speed, cost per token and power per token depend on the model.

Cost per 1M output tokenson Qwen3.5 9B · card over 3 years + power
$1.17
Card, MSRP
$1.01
Power
$0.17
3-year cost of ownership$1,163
Or rent in the cloud$0.14/hr

Renting costs as much as buying after ≈7,300 hours (≈5.0 years at 4 h a day).

Include card price

Each dot is a card; hover or tap it for its numbers, click to open it. Cards without a price for the chosen basis are left out. How we calculate cost per token

Similar GPUs

Cards people weigh against the RTX 2080 Ti, each with the one reason it is worth a look. Speeds on Qwen3.5 9B.

1 GB more and 9% fasterRTX 5070NVIDIA · 12 GB · Blackwell
▲ 9% speed▲ 1 GBabout $840rent $0.14/hr
13 GB more and 52% fasterRTX 3090NVIDIA · 24 GB · Ampere
▲ 52% speed▲ 13 GBabout $1,499rent $0.17/hr
1 GB more, but slowerRTX 3060 12GBNVIDIA · 12 GB · Ampere
▼ 42% speed▲ 1 GBabout $329rent $0.07/hr
5 GB more and 9% fasterRTX 4070 Ti SUPERNVIDIA · 16 GB · Ada Lovelace
▲ 9% speed▲ 5 GBabout $799rent $0.21/hr
5 GB more and 16% fasterRTX 4080NVIDIA · 16 GB · Ada Lovelace
▲ 16% speed▲ 5 GBabout $1,199rent $0.24/hr
5 GB more and 19% fasterRTX 4080 SUPERNVIDIA · 16 GB · Ada Lovelace
▲ 19% speed▲ 5 GBabout $999rent $0.32/hr

RTX 2080 Ti tokens per second by model size

Output speed depends on memory bandwidth (616 GB/s). Grey rows need more than 11 GB.

  1. 9BQwen3.5 9B · Q4_K_M · 6.6 GB
    63.0tok/s
  2. 14BQwen3 14B · IQ4_XS · 10.9 GB tight · IQ4_XS
    46.7tok/s
  3. 27–31BQwen3.8 27B · not at 4-bit
    needs 17.9 GB
  4. 35B MoEQwen3.6 35B A3B · not at 4-bit
    needs 22.0 GB
  5. 70BLlama 3.3 70B Instruct · not at 4-bit
    needs 51.3 GB
  6. 120B MoEGPT OSS 120B · not at 4-bit
    needs 70.8 GB

Estimated speeds: Q4_K_M, 4K context, one user. How we estimate

All 936 models on the RTX 2080 Ti

Featured only

Best quant that fits, ordered by llmrun Score. Fit grades how well it runs here, S best. How ranking works

LLM models compatible with NVIDIA GeForce RTX 2080 Ti — ranked by Score, recency and Fit
ModelVRAMScoreFit
Shieldgemma 2B
Q4_K_M·231.4 t/s
3.6 GBS100, Runs great
Yi 6B Chat
Q4_K_M·98.4 t/s
4.2 GBS91, Runs great
Phi 4 Quantized.w8a8
Q3_K_M·50.9 t/s
10.8 GBC44, Usable
Qwen3.5 27B Claude 4.6 Opus Reasoning Distilled
IQ2_M·39.6 t/s
10.6 GBB50, Decent
Qwen2.5 1.5B
Q4_K_M·312.8 t/s
1.7 GBS100, Runs great
RedPajama INCITE 7B Base
Q4_K_M·86.7 t/s
4.6 GBS89, Runs great
SmolLM 360M
Q4_K_M·667.3 t/s
0.6 GBS100, Runs great
Gemma 1.1 2B IT
Q4_K_M·242.7 t/s
3.5 GBS100, Runs great
SmolLM 360M Instruct
Q4_K_M·667.3 t/s
0.6 GBS100, Runs great
SmolLM2 360M
Q4_K_M·667.3 t/s
0.8 GBS100, Runs great
Codegemma 2B
Q4_K_M·242.7 t/s
3.5 GBS100, Runs great
Fanar 2 27B Instruct
IQ2_XXS·44.0 t/s
10.6 GBB52, Decent
Gemma 3 12B Pt
Q4_K_M·49.8 t/s
10.9 GBC44, Usable
Codegemma 7B IT
Q4_K_M·71.1 t/s
8.2 GBS86, Runs great
Meta Llama 3 8B Instruct
Q4_K_M·74.3 t/s
6.2 GBS86, Runs great

RTX 2080 Ti specs, power and PSU

Specs

Architecture
Turing
VRAM
11 GB GDDR6
Memory bandwidth
616.0 GB/s
FP16 (dense)
56.90 TFLOPS
CUDA cores
4,352
Tensor cores
544
Compute capability
7.5
TDP
250 W
Released
2018-09-27
MSRP
$999

RTX 2080 Ti FAQ

Is the RTX 2080 Ti still worth it in 2026?

At its MSRP of $999 you get 6.3 tok/s per $100 on Qwen3.5 9B, which ranks 3rd of the 4 cards compared here; it is also the 2nd fastest out of 4 cards with 10–11 GB. It runs 7–9B models well; 14B only just fits.

What is the largest model the RTX 2080 Ti can run?

Phi 4 (14.7B) is the largest at 4-bit or better: at Q4_0 it uses 11.0 GB of 11 GB, and runs at about 49.7 tok/s.

Can the RTX 2080 Ti run Qwen3.5 9B?

Yes. Qwen3.5 9B fits at Q4_K_M, using 6.6 GB of 11 GB, and it is the best model that runs well on this card.

Can the RTX 2080 Ti run 27–31B models?

No. Qwen3.8 27B needs 17.9 GB at Q4, more than the 11 GB on this card.

How many tokens per second does the RTX 2080 Ti give?

About 63.0 tok/s on Qwen3.5 9B at Q4_K_M, with the first token after 3.1 s on a 4K-token prompt.

What power supply does the RTX 2080 Ti need?

Plan a 500 W PSU (2× the 250 W TDP).

Is 11 GB enough for coding and agents?

Yes for coding: Llama 3.1 8B Instruct, the only coding-scored model that fits, runs at about 76 tok/s. For agents, not yet: no agent-scored model fits in less than 96 GB.

Is renting cheaper than buying an RTX 2080 Ti?

Rent from $0.14/hr. Renting costs as much as buying after ≈7,300 hours (≈5.0 years at 4 h a day).

Related

Best coding LLMs8 GB VRAM12 GB VRAMGPUs with 11 GB or moreRTX 3080 TiRTX 5070 Ti