llmrun

AMD · RDNA 3 · 24 GB VRAM tier

Best AI Models for AMD Radeon RX 7900 XTX (24 GB)

Runs 27–31B and 35B MoE models well; 70B doesn't fit in 24 GB.

24 GB GDDR6960 GB/s123 TFLOPS FP16355 WRDNA 3 2022
Buyabout $1,130

What LLMs can the RX 7900 XTX run?

#ModelSpeedCoding ScoreQuant · VRAM
1
Qwen3.8 27BAlibabaHow to run →
Q4_K_M · 17.9 GB · 33.1 t/s
33.1 tok/sQ4_K_M · 17.9 GB
2
Gemma 4 31B ITGoogle
Q4_K_M · 21.7 GB · 28.2 t/s
28.2 tok/sQ4_K_M · 21.7 GB
3
Qwen3.6 35B A3BAlibaba
Q4_K_M · 22.0 GB · 76.7 t/s
76.7 tok/sQ4_K_M · 22.0 GB
4
Gemma 4 26B A4B ITGoogle
Q4_K_M · 16.5 GB · 92.9 t/s
92.9 tok/sQ4_K_M · 16.5 GB
5
NVIDIA Nemotron 3.5 Lightning 30B A3B BF16NVIDIA
Q4_K_M · 19.3 GB · 77.5 t/s
77.5 tok/sQ4_K_M · 19.3 GB
See all 327 models on the RX 7900 XTX →Best quant that fits · speeds estimated on this card

About the RX 7900 XTX

The RX 7900 XTX is an entry card in the 24 GB class: it runs 27–31B and 35B MoE models well. It is among the most efficient per watt in its tier, and the best speed per dollar of the cards compared at ≈$1,130. RDNA 3 generation, released 2022.

Speed in 24 GB#4960 GB/s
Per watt#20.26 tok/s per W on Qwen3.5 9B (9B params)
Per dollar#18.0 tok/s per $100 on Qwen3.5 9B (9B params)
GenerationRDNA 3released 2022

RX 7900 XTX strengths and weaknesses

Strengths

  • Runs 27–31B models: Qwen3.8 27B at 33.1 tok/s in 17.9 GB, 6.1 GB left for context
  • Only new 24 GB card here: The other 24 GB cards sell used; this one ≈$1,130 new

Weaknesses

  • 70B models don't fit: Llama 3.3 70B Instruct needs 51.3 GB at Q4 vs 24 GB here
  • Slower on MoE models: Qwen3.6 35B A3B at 76.7 tok/s, slower than Qwen3.5 9B at 90.6
  • Slow prompt reading: A 4K-token prompt on Qwen3.8 27B waits 8.4 s for the first token
  • ROCm, not CUDA: No FP8 weights; MXFP4 only through llama.cpp, not vLLM

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
RX 7900 XTX24 GB
about $1,13090.6
8.0
25.5
RTX 5070 Ti16 GBBest new value
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
RX 907016 GBBest used value
about $65060.4
9.3
27.5
RX 7900 XTX24 GB
about $1,00090.6
9.1
25.5
RTX 5070 Ti16 GB
about $1,10091.6
8.3
30.5
RTX 3090 Ti24 GB
about $1,600103
6.4
22.9
RTX 309024 GB
about $1,50095.7
6.4
27.3
RX 9060 XT 16GB16 GB
about $48030.2
6.3
18.9
RTX 508016 GB
about $1,64098.1
6.0
27.3
RTX 509032 GB
about $6,400183
2.9
31.8

Cost per token: the RX 7900 XTX against 16–32 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.

16 GB24 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 5070 Ti · 16 GB · ≈$1,170 · $0.95 per 1M tokensRX 9070 · 16 GB · ≈$800 · $0.99 per 1M tokensRTX 5080 · 16 GB · ≈$1,690 · $1.25 per 1M tokensRX 9070 XT · 16 GB · ≈$815 · $1.07 per 1M tokensRX 7900 XTX · 24 GB · ≈$1,130 · $0.95 per 1M tokensRX 7900 XTXRTX 5070 TiRX 9070RTX 5090RTX 5080RX 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 5070 Ti · 16 GB · ≈$1,170 · $0.95 per 1M tokensRX 9070 · 16 GB · ≈$800 · $0.99 per 1M tokensRTX 5080 · 16 GB · ≈$1,690 · $1.25 per 1M tokensRX 9070 XT · 16 GB · ≈$815 · $1.07 per 1M tokensRX 7900 XTX · 24 GB · ≈$1,130 · $0.95 per 1M tokensRX 7900 XTXRTX 5070 TiRTX 5090
Cost per 1M output tokenson Qwen3.5 9B · card over 3 years + power
$0.95
Card, new price
$0.79
Power
$0.16
3-year cost of ownership$1,363
Or rent an A100 in the cloud$1.14/hr

Renting costs as much as buying after ≈1,000 hours (≈0.7 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 RX 7900 XTX, each with the one reason it is worth a look. Speeds on Qwen3.5 9B.

6% faster, same 24 GBRTX 3090NVIDIA · 24 GB · Ampere
▲ 6% speedsame 24 GBabout $1,499rent $0.17/hr
8 GB more, 102% faster for $4,246 moreRTX 5090NVIDIA · 32 GB · Blackwell
▲ 102% speed▲ $4,246 more▲ 8 GBabout $5,376rent $0.50/hr
Slower but $540 cheaperRX 9060 XT 16GBAMD · 16 GB · RDNA 4
▼ 67% speed▼ $540 less▼ 8 GBabout $590
Best value near the RX 7900 XTXRTX 5070 TiNVIDIA · 16 GB · Blackwell
same speed▼ 8 GBabout $1,170rent $0.17/hr
8 GB less, 8% faster for $560 moreRTX 5080NVIDIA · 16 GB · Blackwell
▲ 8% speed▲ $560 more▼ 8 GBabout $1,690rent $0.23/hr
14% faster, same 24 GBRTX 3090 TiNVIDIA · 24 GB · Ampere
▲ 14% speedsame 24 GBabout $1,999rent $0.27/hr

RX 7900 XTX tokens per second by model size

Output speed depends on memory bandwidth (960 GB/s). Grey rows need more than 24 GB.

  1. 9BQwen3.5 9B · Q4_K_M · 6.6 GB
    90.6tok/s
  2. 14BQwen3 14B · Q4_K_M · 11.9 GB
    60.6tok/s
  3. 27–31BQwen3.8 27B · Q4_K_M · 17.9 GB
    33.1tok/s
  4. 35B MoEQwen3.6 35B A3B · Q4_K_M · 22.0 GB
    76.7tok/s
  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 327 models on the RX 7900 XTX

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 AMD Radeon RX 7900 XTX — ranked by Score, recency and Fit
ModelVRAMScoreFit
Qwen3.8 27B
Q4_K_M·33.1 t/s
17.9 GBA72, Runs well
Gemma 4 26B A4B IT
Q4_K_M·92.9 t/s
16.5 GBS90, Runs great
Muse Glimmer 30B
Q4_K_M·31.4 t/s
19.6 GBA69, Runs well
NVIDIA Nemotron 3.5 Lightning 30B A3B BF16
Q4_K_M·77.5 t/s
19.3 GBS86, Runs great
GLM 4.7 Flash
Q4_K_M·69.1 t/s
19.9 GBA82, Runs well
Gemma 4 31B IT
Q4_K_M·28.2 t/s
21.7 GBB55, Decent
JEV 27B VL
Q4_K_M·33.1 t/s
17.9 GBA72, Runs well
LFM2.5 24B A2B
Q4_K_M·94.0 t/s
15.0 GBS90, Runs great
GPT OSS 20B
Q4_K_M·107.6 t/s
13.1 GBS92, Runs great
Qwen3.6 35B A3B
Q4_K_M·76.7 t/s
22.0 GBA70, Runs well
MiMo V2.6 Distill Qwen 9B
Q4_K_M·92.8 t/s
6.5 GBS90, Runs great
Ling 3.0 Tiny
Q4_K_M·155.7 t/s
5.5 GBS97, Runs great
Qwen3.5 9B
Q4_K_M·90.6 t/s
6.6 GBS90, Runs great
Qwen3.5 4B
Q4_K_M·176.7 t/s
3.4 GBS99, Runs great
MiniCPM5 2B
Q4_K_M·303.2 t/s
2.5 GBS100, Runs great

RX 7900 XTX specs, power and PSU

Specs

Architecture
RDNA 3
VRAM
24 GB GDDR6
Memory bandwidth
960.0 GB/s
FP16 (dense)
122.80 TFLOPS
Stream processors
6,144
TDP
355 W
Released
2022-12-13
MSRP
$999

RX 7900 XTX FAQ

Is the RX 7900 XTX still worth it in 2026?

At ≈$1,130 you get 8.0 tok/s per $100 on Qwen3.5 9B, which ranks 1st of the 7 cards compared here; it is also the 4th fastest out of 5 cards with 24 GB. It runs 27–31B and 35B MoE models well; 70B doesn't fit in 24 GB.

What is the largest model the RX 7900 XTX can run?

Qwen3.6 35B A3B (36.0B) is the largest at 4-bit or better: at Q4_K_M it uses 22.0 GB of 24 GB, and runs at about 76.7 tok/s.

Can the RX 7900 XTX run Qwen3.8 27B?

Yes. Qwen3.8 27B fits at Q4_K_M, using 17.9 GB of 24 GB, and it is the best model that runs well on this card.

Can the RX 7900 XTX run 70B models?

No. Llama 3.3 70B Instruct needs 51.3 GB at Q4, more than the 24 GB on this card.

How many tokens per second does the RX 7900 XTX give?

About 33.1 tok/s on Qwen3.8 27B at Q4_K_M, with the first token after 8.4 s on a 4K-token prompt.

What power supply does the RX 7900 XTX need?

Plan a 750 W PSU (2× the 355 W TDP).

Is 24 GB enough for coding and agents?

Yes for coding: Qwen3.8 27B, the best of the 9 coding models that fit, runs at about 33 tok/s. For agents, not yet: no agent-scored model fits in less than 96 GB.

Is renting cheaper than buying an RX 7900 XTX?

An A100 (80 GB) rents from $1.14/hr. Renting costs as much as buying after ≈1,000 hours (≈0.7 years at 4 h a day).

Related

Best coding LLMs24 GB VRAM48 GB VRAMGPUs with 24 GB or moreTITAN RTXRX 7900 XT