llmrun

NVIDIA · Blackwell · 12 GB VRAM tier

Best AI Models for NVIDIA GeForce RTX 5070 (12 GB)

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

12 GB GDDR7672 GB/s62 TFLOPS FP16250 WBlackwell 2025
Buyabout $840

What LLMs can the RTX 5070 run?

#ModelSpeedllmrun ScoreQuant · VRAM
1
Qwen3.5 9BAlibabaHow to run →
Q4_K_M · 6.6 GB · 68.7 t/s
68.7 tok/sQ4_K_M · 6.6 GB
2
Qwen3.5 4BAlibaba
Q4_K_M · 3.4 GB · 134 t/s
134 tok/sQ4_K_M · 3.4 GB
3
Gemma 3 12B ITGoogle
Q4_K_M · 10.9 GB · 54.3 t/s
54.3 tok/sQ4_K_M · 10.9 GB
4
Gemma 2 9B ITGoogle
Q4_K_M · 7.3 GB · 71.6 t/s
71.6 tok/sQ4_K_M · 7.3 GB
5
Ling 3.0 TinyInclusion AI
Q4_K_M · 5.5 GB · 289 t/s
289 tok/sQ4_K_M · 5.5 GB
See all 276 models on the RTX 5070 →Best quant that fits · speeds estimated on this card

About the RTX 5070

The RTX 5070 is one of the fastest 12 GB cards: it runs 7–9B models well. It is the most efficient per watt in its tier, and among the best speed per dollar of the cards compared at ≈$840. Blackwell generation, released 2025.

Speed in 12 GB#2672 GB/s
Per watt#10.27 tok/s per W on Qwen3.5 9B (9B params)
Per dollar#28.2 tok/s per $100 on Qwen3.5 9B (9B params)
GenerationBlackwellreleased 2025

RTX 5070 strengths and weaknesses

Strengths

  • Runs 14B models: Qwen3 14B at 46.0 tok/s in 11.9 GB
  • Only new 12 GB card here: The other 12 GB cards sell used; this one ≈$840 new

Weaknesses

  • 27–31B models don't fit: Qwen3.8 27B needs 17.9 GB at Q4 vs 12 GB here
  • Little room for context: Qwen3 14B uses 11.9 of 12 GB, 0.1 GB left
  • Priced well above MSRP: ≈$840 new vs its $549 MSRP (+53%)

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.

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 5070 Ti16 GB
about $1,17091.6
7.8
30.5
RX 907016 GB
about $80060.4
7.6
27.5
RX 9070 XT16 GB
about $81560.4
7.4
19.9
RTX 508016 GB
about $1,69098.1
5.8
27.3
RX 9060 XT 16GB16 GB
about $59030.2
5.1
18.9
RTX 509032 GB
about $5,376183
3.4
31.8
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 4060 Ti 8GB8 GB
about $36229.4
8.1
18.4
RTX 4070 Ti SUPER16 GB
about $90068.7
7.6
24.1
RTX 408016 GB
about $1,10073.3
6.7
22.9
RX 9060 XT 16GB16 GB
about $48030.2
6.3
18.9
RTX 509032 GB
about $6,400183
2.9
31.8

Cost per token: the RTX 5070 against 8–32 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.

8 GB12 GB16 GB32 GBthis card
$0.00$0.50$1.00$1.50$2.00$2.50$0$2,000$4,000$6,000$ per 1M output tokens · lower is betternew price →RTX 5090 · 32 GB · ≈$5,376 · $1.99 per 1M tokensRX 9060 XT 16GB · 16 GB · ≈$590 · $1.46 per 1M tokensRTX 5060 · 8 GB · ≈$460 · $0.77 per 1M tokensRTX 5080 · 16 GB · ≈$1,690 · $1.25 per 1M tokensRTX 5070 Ti · 16 GB · ≈$1,170 · $0.95 per 1M tokensRX 9070 XT · 16 GB · ≈$815 · $1.07 per 1M tokensRX 9070 · 16 GB · ≈$800 · $0.99 per 1M tokensRTX 4060 · 8 GB · ≈$520 · $1.36 per 1M tokensRTX 5070 · 12 GB · ≈$840 · $0.93 per 1M tokensRTX 5070RTX 5060RTX 5090RTX 5080RTX 5070 TiRX 9070 XTRX 9060 XT 16GB
$0.00$0.50$1.00$1.50$2.00$2.50$0$2k$4k$6k$ per 1M output tokens · lower is betternew price →RTX 5090 · 32 GB · ≈$5,376 · $1.99 per 1M tokensRX 9060 XT 16GB · 16 GB · ≈$590 · $1.46 per 1M tokensRTX 5060 · 8 GB · ≈$460 · $0.77 per 1M tokensRTX 5080 · 16 GB · ≈$1,690 · $1.25 per 1M tokensRTX 5070 Ti · 16 GB · ≈$1,170 · $0.95 per 1M tokensRX 9070 XT · 16 GB · ≈$815 · $1.07 per 1M tokensRX 9070 · 16 GB · ≈$800 · $0.99 per 1M tokensRTX 4060 · 8 GB · ≈$520 · $1.36 per 1M tokensRTX 5070 · 12 GB · ≈$840 · $0.93 per 1M tokensRTX 5070RTX 5060RTX 5090
Cost per 1M output tokenson Qwen3.5 9B · card over 3 years + power
$0.93
Card, new price
$0.78
Power
$0.15
3-year cost of ownership$1,004
Or rent in the cloud$0.17/hr

Renting costs as much as buying after ≈5,000 hours (≈3.4 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 5070, each with the one reason it is worth a look. Speeds on Qwen3.5 9B.

4 GB more and 7% fasterRTX 4080NVIDIA · 16 GB · Ada Lovelace
▲ 7% speed▲ 4 GBabout $1,199rent $0.24/hr
20 GB more, 167% faster for $4,536 moreRTX 5090NVIDIA · 32 GB · Blackwell
▲ 167% speed▲ $4,536 more▲ 20 GBabout $5,376rent $0.50/hr
46% slower, same 12 GBRTX 3060 12GBNVIDIA · 12 GB · Ampere
▼ 46% speedsame 12 GBabout $329rent $0.06/hr
Slower but $250 cheaperRX 9060 XT 16GBAMD · 16 GB · RDNA 4
▼ 56% speed▼ $250 less▲ 4 GBabout $590
57% slower, 8 GBRTX 4060 Ti 8GBNVIDIA · 8 GB · Ada Lovelace
▼ 57% speed▼ 4 GBabout $399
12 GB more, 32% faster for $290 moreRX 7900 XTXAMD · 24 GB · RDNA 3
▲ 32% speed▲ $290 more▲ 12 GBabout $1,130

RTX 5070 tokens per second by model size

Output speed depends on memory bandwidth (672 GB/s). Grey rows need more than 12 GB.

  1. 9BQwen3.5 9B · Q4_K_M · 6.6 GB
    68.7tok/s
  2. 14BQwen3 14B · Q4_K_M · 11.9 GB tight · Q4_K_M
    46.0tok/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 956 models on the RTX 5070

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 5070 — ranked by Score, recency and Fit
ModelVRAMScoreFit
DeepSeek OCR
Q4_K_M·390.2 t/s
2.8 GBS100, Runs great
Huihui MiniCPM5 1B Abliterated
Q4_K_M·441.2 t/s
1.3 GBS100, Runs great
Qwythos 9B Claude Mythos 5 1M
Q4_K_M·70.3 t/s
6.5 GBS86, Runs great
Qwen3 1.7B
Q4_K_M·249.6 t/s
3.4 GBS100, Runs great
LFM2.5 1.2B JP 202606
Q4_K_M·408.2 t/s
1.2 GBS100, Runs great
Tiny Random Qwen3
Q4_K_M·1456.0 t/s
0.3 GBS100, Runs great
Unlimited OCR 2
Q4_K_M·388.9 t/s
3.3 GBS100, Runs great
Qwen3 0.6B
Q4_K_M·502.1 t/s
1.7 GBS100, Runs great
Phi Mini MoE Instruct
Q4_K_M·184.8 t/s
5.4 GBS99, Runs great
InternVL3 1B
Q4_K_M·704.5 t/s
2.0 GBS100, Runs great
Tiny Random OPTForCausalLM
BF16·43680.0 t/s
0.0 GBS100, Runs great
Zeta 2.1
Q4_K_M·79.1 t/s
7.4 GBS87, Runs great
CPH Community 7B
Q4_K_M·94.5 t/s
7.1 GBS90, Runs great
Dolphin X1 Trinity Nano
Q4_K_M·215.2 t/s
4.9 GBS100, Runs great
GaMS3 12B Instruct
Q4_K_M·53.9 t/s
8.9 GBA81, Runs well

RTX 5070 specs, power and PSU

Specs

Architecture
Blackwell
VRAM
12 GB GDDR7
Memory bandwidth
672.0 GB/s
FP16 (dense)
61.70 TFLOPS
CUDA cores
6,144
Tensor cores
192
Compute capability
10.0
TDP
250 W
Released
2025-03-05
MSRP
$549

RTX 5070 FAQ

Is the RTX 5070 still worth it in 2026?

At ≈$840 you get 8.2 tok/s per $100 on Qwen3.5 9B, which ranks 2nd of the 8 cards compared here; it is also the 2nd fastest out of 9 cards with 12 GB. It runs 7–9B models well; 14B only just fits.

What is the largest model the RTX 5070 can run?

GPT OSS 20B (20.9B) is the largest at 4-bit or better: at IQ4_XS it uses 11.8 GB of 12 GB, and runs at about 150 tok/s.

Can the RTX 5070 run Qwen3.5 9B?

Yes. Qwen3.5 9B fits at Q4_K_M, using 6.6 GB of 12 GB, and it is the best model that runs well on this card.

Can the RTX 5070 run 27–31B models?

No. Qwen3.8 27B needs 17.9 GB at Q4, more than the 12 GB on this card.

How many tokens per second does the RTX 5070 give?

About 68.7 tok/s on Qwen3.5 9B at Q4_K_M, with the first token after 2.8 s on a 4K-token prompt.

What power supply does the RTX 5070 need?

Plan a 500 W PSU (2× the 250 W TDP).

Is 12 GB enough for coding and agents?

Yes for coding: Llama 3.1 8B Instruct, the only coding-scored model that fits, runs at about 82 tok/s. For agents, not yet: no agent-scored model fits in less than 96 GB.

Is renting cheaper than buying an RTX 5070?

Rent from $0.17/hr. Renting costs as much as buying after ≈5,000 hours (≈3.4 years at 4 h a day).

Related

Best coding LLMs12 GB VRAM16 GB VRAMGPUs with 12 GB or moreRTX 3090 TiRTX 4090