Tesla T10 Price & Used Prices: Which Local LLMs Can It Run?

A single Tesla T10 has 16GB of GDDR6 VRAM. At 4-bit quantization and 8K context it fits a dense model of up to about 23B parameters, or about 13B at 8-bit. MoE models can offload experts to system RAM to run larger ones, such as Qwen3.6 35B A3B (36B).

Not on a single card. At 4-bit with an 8K context a 70B dense model needs about 43GB, more than the Tesla T10's 16GB. It runs if you split the layers across 3 cards with llama.cpp (48GB total).

Among models released in 2026, the largest dense model that fits entirely in VRAM at 4-bit is Gemma 4 12B, at an estimated 39.2 tokens/s. The full list is in the table above.

Estimated from memory bandwidth at 4-bit and 8K context: 8B dense: about 53.4 tokens/s and 14B dense: about 31.7 tokens/s. MoE models read only the active parameters for each token, so they run much faster. Real-world results are usually 70%–100% of the estimate.

With vLLM tensor parallelism at 4-bit and 32K context, two Tesla T10 cards cannot fit any model of 50B or larger released in 2026. At 4-bit with 32K context the minimum is 8 cards, which fits Qwen3.8 Flash Next.

As of Sep 29, 2026: eBay median $495 (5 listings) and Xianyu median $199 (9 listings). Figures are medians of live listings, refreshed every 6 hours.

As of Sep 29, 2026, the eBay median is roughly flat over 7 days, roughly flat over 30 days.

The GeForce RTX 3060 12GB has 12GB of VRAM and 360 GB/s of bandwidth; the Tesla T10 has 16GB and 403 GB/s. At the eBay median, the GeForce RTX 3060 12GB costs about $420, 15.2% less than the Tesla T10 ($495). For 14B dense models at 4-bit, the Tesla T10 is estimated at about 31.7 tokens/s and the GeForce RTX 3060 12GB at about 28.4 tokens/s.