Model vendor
Social media company with AI research
Updated Aug 11, 2026. Data as of 2026-08-11
Default-version specifications and scored-version capability values are labeled separately.
| ME | Muse Spark 1.1 | 88.3 | 40.0% | 1M | Closed weights | Muse Spark 1.1 | |
| ME | Muse Spark 1.2 | 82.3 | 20.0% | 1M | Closed weights | Muse Spark 1.2 | |
| ME | Muse Spark | 73.2 | 60.0% | N/A | Closed weights | Muse Spark | |
| ME | Muse Glimmer-30B | 58.5 | 100.0% | 131.1K | Open weights | Muse Glimmer-30B | |
| ME | Llama 4 Maverick | 24.9 | 80.0% | 1M | Closed weights | Llama 4 Maverick | |
| ME | Llama 3.1 405B Instruct | 24.5 | 100.0% | 128K | Closed weights | Llama 3.1 405B Instruct | |
| ME | Llama 3.3 70B Instruct | 22.1 | 100.0% | 128K | Open weights | Llama 3.3 70B Instruct | |
| ME | Llama 4 Scout | 14.1 | 80.0% | 10M | Closed weights | Llama 4 Scout | |
| ME | Llama 3.1 70B Instruct | 13.2 | 100.0% | 128K | Closed weights | Llama 3.1 70B Instruct | |
| ME | Llama 3.2 90B Instruct | 10.8 | 60.0% | 128K | Closed weights | Llama 3.2 90B Instruct | |
| ME | Llama 3.1 8B Instruct | 0.0 | 100.0% | 131.1K | Closed weights | Llama 3.1 8B Instruct | |
| ME | Llama 3.2 11B Instruct | 0.0 | 60.0% | 128K | Closed weights | Llama 3.2 11B Instruct | |
| ME | Llama 3.2 3B Instruct | 0.0 | 80.0% | 128K | Closed weights | Llama 3.2 3B Instruct |
Each unique model uses its latest scored version.
Leading Meta models by score, price and context window.
Background and model catalog information for Meta.
Social media company with AI research
Model ownership is separate from API hosting. One model can be offered by multiple API providers with different prices.
Catalog data as of 2026-08-11.
Common questions about Meta.
Muse Spark 1.1 currently has the strongest LLMBoard score from this vendor.
The catalog contains 13 independently selectable Meta models and 14 versions.
Not necessarily. Vendor identifies the model developer; API providers are tracked separately.
2 current default versions are marked as open weight.