Best local LLM for 48 GB
For 2× RTX 3090 (48 GB), 2× RTX 4090 (48 GB), RTX 6000 Ada (48 GB), Mac, 64 GB and L40S (48 GB). Method
Behind the frontier10 monthsFirst reached by GPT-5.2-Codex, Dec 2025
Models that fit in 48 GB
Top 10#ModelIndexFormatMemoryCheapest APILowest output price per 1M.
By memory
Index, highest firstMemorySmartest fitIndexFormatBehindAPI equivalent
Method
- Models: ranked models with open weights, an official Hugging Face repository and medium or high confidence. One line per set of weights.
- Memory need: the weights at the format's bits per weight (4-bit 4.85, 8-bit 8.5, 16-bit 16), plus 5% and 1 GB for a short context.
- Format: the most bits that fit. The index is for the full model; a 4-bit copy can score a little lower.
- Shared memory (Mac, DGX Spark, Ryzen AI Max): about two thirds goes to the model up to 36 GB, three quarters above.
- Behind the frontier: months from the first release with the same index or more, among models with medium or high confidence.
- Same by API: the lowest blended price (3 input : 1 output) for that index or more, per 1M tokens, among the same models.