Models › Tessera-27B-B-CJK

Tessera-27B-B-CJK: VRAM needs and speed by GPU

At Q4_K_M, Tessera-27B-B-CJK needs about 18.9 GB of VRAM. No measured runs yet. Speeds below are estimates. Submit a run.

27.32B parameters · license: apache-2.0 · 262,144-token context · GGUF published Oct 18, 2026 · GGUF files on Hugging Face

Memory needed per quantization

QuantFile sizeNeeds (estimate)Smallest card it fits
IQ2_XXS8.9 GB10.4 GB12 GB cards
IQ2_XS9.1 GB10.6 GB12 GB cards
IQ2_S9.7 GB11.2 GB12 GB cards
IQ2_M10.5 GB12.0 GB16 GB cards
Q2_K10.8 GB12.3 GB16 GB cards
IQ3_XXS12.3 GB13.8 GB16 GB cards
Q3_K_S12.7 GB14.2 GB16 GB cards
IQ3_XS12.8 GB14.3 GB16 GB cards
Q3_K_M13.4 GB14.9 GB16 GB cards
Q3_K_L14.1 GB15.6 GB24 GB cards
IQ3_M14.9 GB16.3 GB24 GB cards
IQ4_XS15.5 GB17.0 GB24 GB cards
Q4_016.3 GB17.8 GB24 GB cards
Q4_K_S16.4 GB17.8 GB24 GB cards
IQ4_NL17.4 GB18.9 GB24 GB cards
Q4_K_M17.4 GB18.9 GB24 GB cards
Q4_117.8 GB19.3 GB24 GB cards
Q4_K_L18.8 GB20.3 GB24 GB cards
Q5_K_S19.6 GB21.0 GB24 GB cards
Q5_K_M20.9 GB22.4 GB24 GB cards
Q6_K_S22.9 GB24.3 GB32 GB cards
Q6_K23.9 GB25.3 GB32 GB cards
Q6_K_L25.0 GB26.4 GB32 GB cards
Q8_029.1 GB30.6 GB32 GB cards
BF1654.7 GB56.1 GB96 GB cards

Expected speed on popular hardware

HardwareBest quant that fitsExpected tok/s
NVIDIA RTX 3060IQ2_XXS34.7 estimate
NVIDIA RTX 4060 TiDoes not fit
NVIDIA RTX 3090Q5_K_M31.4 estimate
NVIDIA RTX 4090Q5_K_M34.7 estimate
NVIDIA RTX 5090Q8_039.4 estimate
AMD Radeon RX 7900 XTXQ5_K_M34.9 estimate
Apple M4 Pro (20-core GPU) (64 GB)Q8_08.0 estimate
Apple M3 Max (40-core GPU) (64 GB)Q8_011.6 estimate
AMD Ryzen AI Max+ 395 (Radeon 8060S)Q8_07.1 estimate

Check your own card

Too big for your card?

Run Tessera-27B-B-CJK on a rented 24-80 GB GPU by the hour.

Rent a bigger GPU on RunPod

Affiliate link: if you sign up, RigBench may earn a referral credit. It never affects rankings or numbers.