- Services
Meta released the Llama family of open-weight models, which became the most widely fine-tuned foundation models in the world. Llama models ship under a community license that permits commercial use at most scales.
Visit MetaModels tracked
20
Open weight
14
API only
6
Avg score
53.9
Top benchmark
94.7
AIME 2026
Total HF downloads
6.0M
Context window
2K – 10.0M
First release
Feb 2023
Latest release
Nov 2026
Ranked by composite score across benchmarks, popularity, efficiency, and versatility.
| # | Model | Modality | Score | Params | Released |
|---|---|---|---|---|---|
| 1 | Meta Muse SparkAPI | text | AA77.7 | — | Apr 2026 |
| 2 | Muse ImageAPI | image | AA72.8 | — | Jul 2026 |
| 3 | Muse Spark 1.1API | text | AA71.8 | — | Jul 2026 |
| 4 | text | BB65.6 | 30B | Aug 2026 | |
| 5 | text | BB62.7 | 65B | Feb 2023 | |
| 6 | Muse WonderAPI | text | BB57.6 | — | Oct 2026 |
| 7 | text | BB56.1 | 70B | Dec 2024 | |
| 8 | text | BB55.3 | 8B | Jul 2024 | |
| 9 | Muse EmberAPI | text | CC53.2 | — | Nov 2026 |
| 10 | text | CC52.5 | 816B | Oct 2026 |
The strongest model from this provider in each modality. Click a card to open the model page.
Text
Meta Muse Spark
Image
Muse Image
Video
Muse Video
How this provider’s catalog splits across text, image, video, audio, embedding, and decision models.
When each model shipped, newest first.
Nov 9
Oct 15
Oct 12
Aug 10
Jul 9
Jul 7
Jul 7
Apr 7
Apr 4
Apr 4
Jul 22
Jul 22
Jul 22
Apr 17
Apr 17
Jul 17
Jul 17
Jul 17
Feb 23
Composite grades across this provider’s catalog. Higher is better, blending benchmarks, popularity, and efficiency.
Models with downloadable weights, ranked by composite score.
| # | Model | Modality | Score | Params | Released |
|---|---|---|---|---|---|
| 1 | text | BB65.6 | 30B | Aug 2026 | |
| 2 | text | BB62.7 | 65B | Feb 2023 | |
| 3 | text | BB56.1 | 70B | Dec 2024 | |
| 4 | text | BB55.3 | 8B | Jul 2024 | |
| 5 | text | CC52.5 | 816B | Oct 2026 | |
| 6 | text | CC51.2 | 8B | Apr 2024 | |
| 7 | text | CC50.4 | 400B | Apr 2025 | |
| 8 | text | CC49.8 | 70B | Jul 2024 | |
| 9 | text | CC47.8 | 109B | Apr 2025 | |
| 10 | text | CC43.6 | 70B | Apr 2024 | |
| 11 | text | CC43.0 | 13B | Jul 2023 | |
| 12 | text | CC41.9 | 405B | Jul 2024 | |
| 13 | text | CC40.4 | 7B | Jul 2023 | |
| 14 | text | DD38.0 | 70B | Jul 2023 |
Closed-source models accessed through Meta’s API.
API pricing is per million tokens unless noted otherwise.
| # | Model | Modality | Score | Params | Released |
|---|---|---|---|---|---|
| 1 | Meta Muse SparkAPI | text | AA77.7 | — | Apr 2026 |
| 2 | Muse ImageAPI | image | AA72.8 | — | Jul 2026 |
| 3 | Muse Spark 1.1API | text | AA71.8 | — | Jul 2026 |
| 4 | Muse WonderAPI | text | BB57.6 | — | Oct 2026 |
| 5 | Muse EmberAPI | text | CC53.2 | — | Nov 2026 |
| 6 | Muse VideoAPI | video | CC46.5 | — | Jul 2026 |
Each family is a model series. Browse all releases in a family with their own leaderboard.
Meta ships 20 models that we track. Of those, 14 are open weight and 6 are accessed through their API.
It depends on the model. See the open-weight section above for models you can run locally, and the API-only section for closed models priced per token.
By composite score, Muse Glimmer-30B is currently the top Meta model in our directory. Composite score blends benchmarks, popularity, efficiency, and versatility.
Full Directory
Open the full directory to filter by hardware, capability, license, and benchmark score.
Open Directory
See every release in a family side by side, with a timeline and a leaderboard. Useful when you have already picked a series and want the right size.
Benchmark data sourced from: Hugging Face, Arena.ai, Artificial Analysis.
