mixture-of-experts

Everything tagged mixture-of-experts.

Share your story Subscribe

Featured

News

Recursive releases Flint open-weight coding model for local hardware

DP News TeamOct 21, 2026

News

AWS releases MoE inference changes for SageMaker LMI and vLLM

Oct 17, 2026

News

Reflection AI releases Beam weights under Apache 2.0

Oct 17, 2026

Latest

News
Microsoft completes pre-training of 2.8-trillion-parameter MAI-Base-2
Microsoft says it completed pre-training of the 2.8-trillion-parameter MAI-Base-2, while post-training of the mixture-of-experts model remains…
DP News Team · 2w ago
News
Meta releases 816B-parameter Muse Spark Open model
Meta released downloadable BF16 and NVFP4 weights for Muse Spark Open, an 816-billion-parameter multimodal mixture-of-experts model. Apple's same-day…
DP News Team · 2w ago
News
Liquid AI releases 24B mixture-of-experts model for on-device agents
Liquid AI has released LFM2.5-24B-A2B, an open-weight reasoning model that activates 2.45 billion of its 24 billion parameters per token and targets…
DP News Team · 2w ago
News
Mistral launches Large 4 in public preview
Mistral AI launched Large 4 in public preview, documenting a multimodal mixture-of-experts design, 49 billion active parameters and API access.
DP News Team · 3w ago
News
AllenAI releases Olmo-core 3 for open mixture-of-experts training
AllenAI released Olmo-core 3, an open training stack that distributes mixture-of-experts routing and optimizer work and supports selective MXFP8…
DP News Team · 4w ago
News
Xiaomi releases MiMo-V2.6-Pro-RL open-weight multimodal model
Xiaomi MiMo released downloadable weights for MiMo-V2.6-Pro-RL, which it describes as a sparse mixture-of-experts model for text, images, video and…
DP News Team · Sep 22
News
Cohere releases North Small Translate, an open-weight translation model
Cohere released North Small Translate with API access, non-commercial FP8 weights on Hugging Face, and a Model Vault path for private deployment.
DP News Team · Sep 10
News
MBZUAI launches six K2 Horizon models spanning 0.9B to 375B parameters
MBZUAI’s Institute of Foundation Models released six K2 Horizon models spanning edge to enterprise uses, but its repository inventories show some…
DP News Team · Sep 9
News
Thinking Machines releases Inkling-Small, a leaner open-weight sibling of its flagship
Thinking Machines Lab released Inkling-Small on July 30, 2026, an open-weight model scoring 40 on Artificial Analysis's Intelligence Index at a third…
DP News Team · Jul 30
News
Thinking Machines Lab releases Inkling, its first open-weight multimodal model
Thinking Machines Lab, founded by former OpenAI CTO Mira Murati, released Inkling on July 15, 2026, a 975-billion-parameter open-weight multimodal…
DP News Team · Jul 16
News
NVIDIA's compressed Puzzle-75B and audio-text Audex Nemotron models climb Hugging Face trending list
Two NVIDIA Nemotron models published in early July reached Hugging Face's trending list, led by Puzzle-75B, which NVIDIA says runs 2.03x faster than…
Dmytro Spodarets · Jul 11
News
Tencent's open-weight Hy3 model surges up Hugging Face's trending charts
Tencent's Hy3, an open-weight 295-billion-parameter Mixture-of-Experts model released July 6, surged up Hugging Face's trending charts on July 9,…
Dmytro Spodarets · Jul 9

Featured

Recursive releases Flint open-weight coding model for local hardware

AWS releases MoE inference changes for SageMaker LMI and vLLM

Reflection AI releases Beam weights under Apache 2.0

Latest

Microsoft completes pre-training of 2.8-trillion-parameter MAI-Base-2

Meta releases 816B-parameter Muse Spark Open model

Liquid AI releases 24B mixture-of-experts model for on-device agents

Mistral launches Large 4 in public preview

AllenAI releases Olmo-core 3 for open mixture-of-experts training

Xiaomi releases MiMo-V2.6-Pro-RL open-weight multimodal model

Cohere releases North Small Translate, an open-weight translation model

MBZUAI launches six K2 Horizon models spanning 0.9B to 375B parameters

Thinking Machines releases Inkling-Small, a leaner open-weight sibling of its flagship

Thinking Machines Lab releases Inkling, its first open-weight multimodal model

NVIDIA's compressed Puzzle-75B and audio-text Audex Nemotron models climb Hugging Face trending list

Tencent's open-weight Hy3 model surges up Hugging Face's trending charts