whatmodel.sh / LLaMA 33B
LLaMA 33B
↗ share
+ contribute
LLaMA 33B is an open-weights language model from Meta, released February 2023. Its published size is 32.5B parameters. Licence: llama-1-noncommercial-research.
Catalogue identity notes
Original LLaMA (v1) base model (Touvron et al. 2023), released by Meta to approved researchers under a non-commercial licence; no official public download exists, so no run command is listed (community re-uploads such as huggyllama/* are unofficial). Meta's paper calls it 33B; the weights were distributed (and are widely mirrored) as 'llama-30b'.
$ how it compares
The LLMs ranked either side of LLaMA 33B, out of 828 with a score, then more from Meta.
| # | model | score | where | ||
|---|---|---|---|---|---|
| 680 |  Granite 3.1 2B Instruct | 36±1 | local | ||
| 681 |  EXAONE 3.5 2.4B Instruct | 36±2 | local | ||
| 682 | LLaMA 33B | 36±1 | comparison only | ||
| 683 | Code Llama 7B Instruct | 36±1 | local | ||
| 684 |  LFM2 700M | 36±2 | local | ||
| more from Meta | |||||
| 24 | Muse Spark 1.3 | 77±1 | api $2.00/M | ||
| 40 | Muse Spark 1.2 | 75±1 | api $2.00/M | ||
| 43 | Muse Spark 1.1 | 75±1 | api $2.00/M | ||
$ what it's good at
coding▁▁▁▃▅▇█▇▅▄▃▁ #495 of 524 · 21% of its score · 2 tests
Beats Qwen3.5 2B, Molmo 7B-D (0924), DeepSeek R1 Distill Qwen 1.5B here, though they rank higher overall. Trails Granite 3.0 1B-A400M Instruct, Llama 3.2 1B Instruct, SmolLM2 1.7B Instruct here, though they rank lower overall.
math▁▂▃▅▇▅▅█▄▁▁▁ #424 of 462 · 15% of its score · 2 tests
Beats Yi 34B Chat, Llama 3 8B Instruct, Command R (03-2024) here, though they rank higher overall. Trails Gemma 3 1B, OLMo 2 1B Instruct (0425), Stable Code 3B here, though they rank lower overall.
other tests13 not in a task · 64% of its score
| test | result | vs the others in its run | counts |
|---|---|---|---|
| ARC Challenge +1 | 54.4% ✎ | ▄···▄···█▄·▄ beats 2/5 | 12% |
| WinoGrande +1 | 76.2% ✎ | █··█·█·██··█ beats 3/5 | 8.2% |
| MMLU +1 | 56.8% ✎ | ▄····▄▄·▄··█ beats 2/5 | 8.1% |
| HellaSwag +1 | 83.7% ✎ | █···██···███ beats 3/5 | 6.0% |
| TriviaQA | 68.5% ✎ | █···██··█·██ beats 3/5 | 4.7% |
| Natural Questions (closed book) | 24.1% ✎ | █·····▄▄▄··▄ beats 3/5 | 4.6% |
| Factual knowledge recall, Premium2k exact match (zero-shot) | 27.3% ✎ | ▅····▃▅·██·█ beats 9/13 | 4.5% |
| ARC Easy +1 | 79.6% ✎ | ▄···▄···█▄·▄ beats 2/5 | 3.8% |
5 more tests
| test | result | vs the others in its run | counts |
|---|---|---|---|
| BoolQ lab | 83.1% ✎ | ▁▄▅▆▂▆▄▄█▃▄▄ beats 46/58 | 3.1% |
| Factual knowledge recall, Premium2k exact match (10-shot) | 48.9% ✎ | ▃▅▃··▅▃·█▃▃▅ beats 11/13 | 3.1% |
| OpenBookQA lab | 58.6% ✎ | █▁▁▁▄▃▃▇▃▁▄▂ beats 53/65 | 2.7% |
| PIQA +1 | 82.2% ✎ | ▄····▄··█▄·▄ beats 3/5 | 2.4% |
| Social IQA lab | 50.4% ✎ | ▂·▂██▅▅▁▂▃▆▁ beats 32/51 | 1.4% |
$ where to run it
No verified API or download recorded yet · add one
$ how the score was worked out · Mixtral of Experts — base-mod… 52%, LLaMA: Open and Efficient Fou… 40%, Towards a Holistic Evaluation… 8%
Each result implies a score on its own, after adjusting for how hard its test is and who ran it. The score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its counts column above; here it's added up by who ran the tests. How it works
| source | results | share of the score |
|---|---|---|
 Mixtral of Experts — base-model evaluations | 12 | 52% |
| LLaMA: Open and Efficient Foundation Language Models — authors' own runs · lab-reported | 13 | 40% |
| Towards a Holistic Evaluation of LLMs on Factual Knowledge Recall (arXiv:2404.16164v1) | 2 | 8% |
| prior · pull toward the middle | 0% | |
| weighted average | = 36.1 |
$ improve this page
Use a short GitHub form or edit the data through a pull request. Every number needs a source; here's which ones we can use.
data updated · sources and licences · json