RTX 3090 24GB for local AI in the UK
Ranks #17 of 23 for 7-9B models on value (£7 per token/s at £971 used).
Key facts
- Cheapest UK price
- £971 used
- Memory
- 24 GB
- Memory bandwidth
- 936 GB/s
- Largest model it can run
- ~36B parameters at 4-bit (estimate: ~0.6 GB per billion parameters + ~2 GB overhead)
- Shop
- eBay UK (used)
- £971 used: median of 256 UK eBay sold listings, 90 days, checked date unknown (source)
Disclosure: eBay "Shop" links are affiliate links (marked rel="sponsored"): if you buy, we may earn a small commission at no extra cost to you. Rankings are never affected by commission.
Speed and value by model size
| Size class | Fits at 4-bit? | Speed | £ per tok/s | Value rank | Reference model |
|---|---|---|---|---|---|
| 7-9B models | Yes | 143.5 words/sec* from 35 tests other models: 88.0 (only 2 tests ⚠) | £7 | #17 of 23 | Llama 2 7B Q4_0 |
| 12-14B models | Yes | no verified data yet other models: 78.2 (from 3 tests) | – | not ranked | Gemma 4 12B Q4_0 |
| 27-35B models | Yes | 42.7 words/sec* only 1 test ⚠ other models: 38.6 (from 14 tests) | £23 | low sample (n=1) ⚠, not ranked | Qwen 3.6 27B Q4_K_M |
| 70B+ models | No (needs ~42 GB+) | – | – | – | no reference model yet |
Speed is the median single-user decode speed running the like-for-like reference model for that class, in tokens per second (*a token is roughly ¾ of a word, so words/sec is a rough guide). Ranks use the cheapest UK price and need at least 3 test runs; ⚠ = low sample, shown but not ranked. "Fits" means the model fits fully in memory at ~4-bit. How we measure.
Other models measured
- qwen2.5-coder Q4_K_M (7.6B, 7-9B class): 54.5 tok/s only 2 tests ⚠ · run
- llama3.1 Q4_K_M (8.0B, 7-9B class): 136.2 tok/s only 1 test ⚠ · run
- gemma3 Q4_K_M (11.8B, 12-14B class): 78.7 tok/s only 1 test ⚠ · run
- qwen2.5-coder-14b Q4_K_M (14.0B, 12-14B class): 78.2 tok/s only 1 test ⚠ · run
- deepseek-coder-v2 Q4_0 (15.7B, 12-14B class): 189.5 tok/s only 1 test ⚠ · run
- gpt-oss-20b Q4_K_M (20.0B): 207.2 tok/s only 1 test ⚠ · run
- Qwen 3.8 27B Q4_K_M (27.0B, 27-35B class): 37.6 tok/s only 1 test ⚠ · run
- Qwen 35 27B Q5_K_M (27.0B, 27-35B class): 38.8 tok/s from 12 tests · run
- glm-4.7-flash Q4_K_M (29.9B, 27-35B class): 134.6 tok/s only 1 test ⚠ · run
- qwen3-coder-30b-a3b Q4_K_M (30.0B, 27-35B class): 199.7 tok/s only 1 test ⚠ · run
- Qwen 3 32B Q4_K_M (32.0B, 27-35B class): 38.5 tok/s only 1 test ⚠ · run
- qwen35moe-35b.a3b Q5_K_M (35.0B, 27-35B class): 150.4 tok/s from 18 tests · run
- nvidia-nemotron-3.5-bolt-74b-a9b-24gb (74.0B, 70B+ class): 87.8 tok/s only 1 test ⚠ · run
- nvidia-nemotron-3.5-bolt-74b-a9b-24gb-ctx8k (74.0B, 70B+ class): 80.3 tok/s only 1 test ⚠ · run
Shown for information only: these don't affect rank. Mixture-of-experts models (names with "a3b", "moe" and similar) run much faster than their size suggests.
Similar price alternatives
- RTX 4070 Ti Super 16GB £784 used · 16 GB · 134.3 tok/s on 7-9B · #15 of 23 for 7-9B
- RTX 4080 Super 16GB £825 used · 16 GB · 147.9 tok/s on 7-9B · #13 of 23 for 7-9B
- RTX 5070 Ti 16GB £921 used · 16 GB · 168.2 tok/s on 7-9B · #12 of 23 for 7-9B
- RTX 3090 Ti 24GB £1,020 used · 24 GB · 155.6 tok/s on 7-9B · #16 of 23 for 7-9B
See the RTX 3090 24GB on the leaderboard → · Best under £1,000 · All cards · Beginner's guide · Methodology
Weekly UK value moves, by email
Price drops, new benchmark results, best buys under £500 / £1,000 / £2,000.
Check your inbox to confirm: we use double opt-in, so you're only added after clicking the link in the confirmation email. Free; unsubscribe any time. Sent via Mailchimp (see their privacy notice).