Local LLM · Qwen4-27B · Used GPUs in Japan · 2026-09
This is a tool for buying a second-hand GPU machine to run local LLMs — uncensored models included — at home. It searches Mercari, Yahoo! Auctions, Yahoo! Flea Market and Rakuma in one go, then lets you paste any listing and tells you whether it's cheap or expensive for what's inside. It covers the V100 32GB, modified RTX 3080 20GB, RTX 3090 / 3090 Ti, 4090, 5090 and the EVO-X2.
The marketplaces are Japanese and ship domestically; outside Japan you'll need a proxy-buying service. Builds up to about ¥500k are compared in the second half of The ¥6M local-LLM rig.
The short answer (September 2026, waiting for Qwen4-27B)
Alibaba announced Qwen4-27B on 22 September; the weights and specs aren't out yet. Uncensored builds are expected to follow soon after release. Uncensoring costs a little intelligence, so you'll want to run it at a higher quantization (Q5–Q6), which makes 32GB or more the comfortable line. From there, pick by what you'll use it for:
Pick a keyword and a price range to get links to each site's results, cheapest first and only items still for sale. "Sold prices" shows what things actually went for. Keep the keywords in Japanese — that's how sellers write them.
These are plain links to each site's own search page. This site doesn't fetch or store anything from them.
Copy the title, description and price from a listing, paste it here and hit "Read listing". Fix any field it gets wrong by hand. Japanese listings work as-is, and it handles both bare cards and complete PCs.
Estimated value = the GPU(s). For a complete PC, add the platform (CPU and board), RAM, SSD, PSU and case. The EVO-X2 is priced as a whole machine. A value-to-price ratio of 1.20 or more is a great deal, 1.00–1.20 is around market, and below that is overpriced. GPU and EVO-X2 figures are September 2026 Yahoo! Auctions medians (the 3080 20GB is a rough guide — there are very few sales); the rest are typical private-sale prices for used parts. Change them when the market moves.
How fast an LLM writes is set mostly by memory bandwidth, not raw compute. Below is the simple ceiling for a 27B model at Q4 (about 17 GB — bandwidth ÷ 17 GB) next to September 2026 second-hand prices.
| Hardware | VRAM | Bandwidth | 27B Q4 ceiling | Used price | Per GB | Best for |
|---|---|---|---|---|---|---|
| Tesla V100 32GB | 32GB | 900GB/s | ~53 tok/s | ¥108,000 | ~¥3,400 | Chat. The cheapest 32GB |
| RTX 3080 20GB (modified) | 20GB | 760GB/s | ~45 tok/s | ¥69,000 (rough) | ~¥3,500 | Two cards for 40GB |
| RTX 3090 | 24GB | 936GB/s | ~55 tok/s | ¥200,000 | ~¥8,300 | All-rounder: LLMs and video |
| RTX 4090 | 24GB | 1,008GB/s | ~59 tok/s | ¥428,000 | ~¥17,800 | Overpriced for LLMs alone |
| RTX 5090 | 32GB | 1,792GB/s | ~105 tok/s | ¥910,000 | ~¥28,400 | Fastest, if money is no object |
| EVO-X2 64GB | 64GB (shared) | 256GB/s | ~15 tok/s | ¥249,000 (whole box) | ~¥3,900 | No CUDA, big MoE models, silence |
Not sure yet? You can rent a comparable GPU by the hour and check whether the model you want is good enough for your use before buying — and avoid a machine you end up barely using.
Qwen4-27B's specs aren't public yet. This table assumes it's a dense 27B model like the recent Qwen 27B releases and works out the weight size at each quantization from bits per parameter. Conversation history (the KV cache) needs a few GB on top.
| Quant | Size | 24GB 3090 | 32GB V100, 5090 | 40–48GB two cards | 64GB EVO-X2 |
|---|---|---|---|---|---|
| Q4_K_M | ~16.5GB | Comfortable | Comfortable | Comfortable | Fits (slow) |
| Q5_K_M | ~19GB | Fits (shorter context) | Comfortable | Comfortable | Fits (slow) |
| Q6_K | ~22GB | Just barely | Fits | Comfortable | Fits (slow) |
| Q8_0 | ~29GB | Doesn't fit | Just barely | Comfortable | Fits (slow) |
Uncensored models (made by abliteration or similar pruning) lose a little intelligence compared with the original. You don't want to lose more to heavy quantization, so 32GB for Q5–Q6 is the comfortable line, and 40–48GB across two cards if you want Q8 with long context. A single 24GB 3090 still runs Q4–Q5 well. If Qwen4-27B turns out to be a mixture-of-experts model rather than dense, its total size will grow, and the EVO-X2's 64–128GB comes into its own.
Yahoo! Auctions sold listings, excluding junk, untested units, multi-card bundles, and absurdly cheap box-only sales. These are prices things actually sold for, not asking prices.
| What | Sales · period | Median | Range |
|---|---|---|---|
| Tesla V100 32GB (mostly with fan) | 45 2026-09-04 – 09-24 | ¥108,000 | ¥108,000 – 125,000 |
| HP Z440 + V100 32GB bundle | 2 2026-09 | — | ¥128,000 and ¥178,000 |
| RTX 3080 20GB (modified) | 2 2026-04 – 06 | ¥69,000 | A junk unit went for ¥36,000 |
| RTX 3090 card | 42 2026-08-29 – 09-23 | ¥200,000 | ¥185,000 – 210,000 |
| Desktop with 3090 / 3090 Ti | 16 2026-07-12 – 09-21 | ¥239,000 | ¥202,000 – 261,000 |
| EVO-X2 64GB | 8 2026-03-29 – 09-02 | ¥249,000 | ¥233,000 – 268,000 |
| EVO-X2 128GB | 3 in Sept / 13 in Apr–Aug | ¥500,000 / ¥400,000 | Up about 25% in September |
| RTX 4090 card | 32 2026-08-01 – 09-23 | ¥428,000 | ¥410,000 – 440,000 |
| Desktop with 4090 | 15 2026-07-17 – 09-22 | ¥520,000 | ¥493,000 – 574,000 |
| RTX 5090 card (September) | 18 2026-09-01 – 09-23 | ¥910,000 | ¥745,000 in late August — up 20% in a month |
Since early September, one seller has been steadily listing V100s "with fan and cable, tested" in the low ¥100k range. With bare 3090s this expensive, a complete 3090 PC gets you the rest of the machine for about ¥40k more. Mercari only renders results in a browser, so it isn't in the numbers — use the links above to check it directly.
None of the links on this English page are affiliate links. The Japanese page links to store searches on Amazon and Rakuten with affiliate tags; the marketplace links are never affiliate links, and verdicts don't depend on referral fees.