Best value for memory
Ranked models by Overall score per GB of estimated memory need (the model file size plus about 20% for the context of a single conversation), highest first. A small model with a decent score can win this list without being the best model.
222 of 227 ranked models have a file size on record (so an estimated memory need) and are listed here.
Read the memory figure as an estimate, not a guarantee. The memory a model needs is estimated from its file: the weights plus about 20% for the context of a single conversation. It is not a measurement, and a long conversation or several at once needs more. A model with no file size on record is not counted as fitting any amount of memory. Speeds, and the GPU memory shown on the speed page, were measured on this benchmark's server, not on yours. That server kept 1 to 8 answers in flight at a time and loaded a context window of 16,384 tokens for most models; 42 of the models ranked ran at a shorter context. Speeds were measured on one laptop and are comparable only between models on the same backend. See the settings and the memory method.
Overall points per GB
Smollm2 135M Instruct gives the most Overall score for its memory: 159.66 points per GB, an Overall of 20.1 (95% interval 16.2–24.6) at about 0.1 GB (file size plus 20%).
| Place | Model | Overall per est. GB | Overall | Rank range | Decode (tok/s) | Memory needed (GB, est.) | Size class |
|---|---|---|---|---|---|---|---|
| 1 | Smollm2 135M Instruct | 159.66 | 20.116.2–24.6 | 220–227 | 179.9ROCm 7.1 | 0.1 | Tiny (under 3B) |
| 2 | Granite4 350M H | 103.80 | 45.640.7–50.7 | 184–208 | 55.8ROCm 7.1 | 0.4 | Tiny (under 3B) |
| 3 | Qwen2.5 0.5B Instruct | 90.69 | 43.337.4–49.4 | 186–212 | 99.3ROCm 7.1 | 0.5 | Tiny (under 3B) |
| 4 | Smollm2 360M Instruct | 86.33 | 28.123.4–32.9 | 212–223 | 275.3ROCm 7.1 | 0.3 | Tiny (under 3B) |
| 5 | Lfm2.5 Thinking 1.2B | 84.71 | 74.369.5–79.1 | 84–148 | 205.0ROCm 7.1 | 0.9 | Tiny (under 3B) |
A higher place is not always a clear win. When two models have overlapping rank ranges, this battery cannot tell them apart; see how sure a score is. A dash means not measured or not scored, which is different from 0. Speeds were measured on one laptop and are comparable only between models on the same backend.
More: all the best-model pages, the hardware picker, speed and memory for every model.