News
ReleasesOVHcloud unveils Olala mini, its first in-house LLMChinaA 9B model beats a 1T model on FinQA, trained with Crusoe Managed RLReleasesThe internal metaphysics of an LLMReleasesMapping Language Model Internal Representations to Physical Laws and Scientific ReasoningResearchCan China keep rogue AI in check? Meet the start-up testing the top modelsFrontierMicrosoft aims to challenge OpenAI, Anthropic with cheaper, specialized AI modelsOpen modelsMicrosoft rolls out new first-party models and updates to agentic AI continuous improvementFrontierMicrosoft unveils AI models for coding, cybersecurity as it takes on Anthropic, OpenAIFrontierGoogle is making it easier to opt out of Gemini AI model training – here’s howFrontierOpenAI Unveils New Astra AI Model After Scrapping Prior ReleaseReleasesOVHcloud unveils Olala mini, its first in-house LLMChinaA 9B model beats a 1T model on FinQA, trained with Crusoe Managed RLReleasesThe internal metaphysics of an LLMReleasesMapping Language Model Internal Representations to Physical Laws and Scientific ReasoningResearchCan China keep rogue AI in check? Meet the start-up testing the top modelsFrontierMicrosoft aims to challenge OpenAI, Anthropic with cheaper, specialized AI modelsOpen modelsMicrosoft rolls out new first-party models and updates to agentic AI continuous improvementFrontierMicrosoft unveils AI models for coding, cybersecurity as it takes on Anthropic, OpenAIFrontierGoogle is making it easier to opt out of Gemini AI model training – here’s howFrontierOpenAI Unveils New Astra AI Model After Scrapping Prior Release
EveryLLM

Hybrid designs

Jamba and others that mix attention with state-space layers. 31 models on EveryLLM so far.

Hybrid models combine a few attention layers with many cheaper layers of another kind, usually state-space (Mamba) layers or forms of linear attention. Attention gives exact recall over the context, while the cheaper layers carry most of the load, cutting the memory needed for long inputs and speeding up generation. Since 2024 hybrids have moved from experiments into production models from AI21, Nvidia, IBM and Alibaba.

Jamba and after

AI21 Labs in Tel Aviv released Jamba in March 2024, mixing transformer and Mamba layers with a mixture of experts: 52 billion parameters in total, 12 billion active, and a context window of 256,000 tokens. Nvidia's Nemotron Nano 2 (August 2025) has 62 layers, of which only 6 use self-attention, and Nvidia reported up to six times the throughput of a similar-sized transformer on reasoning tasks.

Mainstream releases

IBM's Granite 4.0 family (October 2025) pairs Mamba-2 and transformer layers at roughly nine to one, and IBM reported over 70% less memory for long inputs. Alibaba's Qwen3-Next (September 2025) alternates three Gated DeltaNet layers, a form of linear attention, with one gated attention layer, in an 80-billion-parameter mixture of experts with 3 billion active. Such designs aim at long documents and agents that run for many steps.

Famous hybrid models

Further reading

imgwavvidNemotron 🇺🇸 United StatesNemotron Nvidia family of language and multimodal models for enterprise and reasoningGeneral · Conditional licence · 2023 Evo 🇺🇸 United StatesEvo Open genomic language models that read and write DNA, RNA and protein sequencesDomain · Fully open · 2024 RecurrentGemma 🇺🇸 United StatesRecurrentGemma Open language models using Google's Griffin recurrent-plus-attention architectureGeneral · Open weights · 2024 MiMo-V2-Flash 🇨🇳 ChinaMiMo-V2-Flash High-speed MoE language model for reasoning and agentsGeneral · Open weights Snowflake Arctic 🇺🇸 United StatesSnowflake Arctic Open enterprise language model with hybrid MoE designGeneral · Open weights · 2024 fnNemotron Nano 2 🇺🇸 United StatesNemotron Nano 2 A 9B hybrid Mamba-Transformer model for reasoning and chat, from NVIDIA.General · Open weights · 2025 fnNemotron 3 Super 🇺🇸 United StatesNemotron 3 Super NVIDIA's 120B hybrid MoE language model with 12B active parameters and 1M-token contextGeneral · Conditional licence · 2026 imgBLIP-2 WorldBLIP-2 Efficient vision-language model using frozen encodersMultimodal · Research only · 2023 Zamba WorldZamba A 7-billion parameter hybrid language model combining Mamba and transformer layers.General · Open weights · 2024 fnLFM2 🇺🇸 United StatesLFM2 Hybrid language models for edge and on-device useSmall · Open weights PLaMo 2 🇯🇵 JapanPLaMo 2 Japanese-focused large language model series by Preferred NetworksGeneral · Open weights · 2025 fnQwen3-Next-80B-A3B 🇨🇳 ChinaQwen3-Next-80B-A3B An 80-billion parameter hybrid model with ultra-long context support.General · Open weights imgfnQwen3.5-397B-A17B 🇨🇳 ChinaQwen3.5-397B-A17B A large multimodal language model with sparse activation.General · Open weights imgfnQwen3.6-35B-A3B 🇨🇳 ChinaQwen3.6-35B-A3B Open-weight multimodal coding model with hybrid architectureCode · Open weights · 2026 fnEXAONE 4.0 🇰🇷 South KoreaEXAONE 4.0 Unified language model with reasoning and non-reasoning modesGeneral · Conditional licence · 2025 fnSolar Open 2 🇰🇷 South KoreaSolar Open 2 250B parameter open-weight model for agentic tasksGeneral · Open weights · 2026 Qwen3.8-2.4T-A95B 🇨🇳 ChinaQwen3.8-2.4T-A95B Open-weight reasoning model with 2.4 trillion parametersReasoning · Open weights · 2026 Falcon-H1R-7B 🇦🇪 United Arab EmiratesFalcon-H1R-7B A 7-billion parameter reasoning model with hybrid architectureReasoning · Open weights · 2026 Falcon-H1 🇦🇪 United Arab EmiratesFalcon-H1 Hybrid transformer and state-space model series by TIIGeneral · Open weights · 2025 Kimi Linear 🇨🇳 ChinaKimi Linear Hybrid linear attention model with 48 billion parametersGeneral · Open weights · 2025 MiniMax-Text-01 🇨🇳 ChinaMiniMax-Text-01 A large language model with a four million token context window.General · Open weights · 2025 fnMiniMax-M1 🇨🇳 ChinaMiniMax-M1 Open-weight reasoning model with hybrid attention and long contextReasoning · Open weights · 2025 ERNIE 3.0 🇨🇳 ChinaERNIE 3.0 Knowledge-enhanced pre-training model for language tasksGeneral · Research only · 2021 Hunyuan-TurboS 🇨🇳 ChinaHunyuan-TurboS Hybrid Mamba-Transformer MoE model by TencentGeneral · API only · 2025 MiMo-V2.5-Pro 🇨🇳 ChinaMiMo-V2.5-Pro Open-source MoE language model with 1M token contextGeneral · Open weights imgvidNemotron 3.5 Super VL 🇺🇸 United StatesNemotron 3.5 Super VL NVIDIA's open-weight vision-language model with 120B parameters and 1M-token contextMultimodal · Conditional licence · 2026 Jamba2 Mini WorldJamba2 Mini A small open-weights SSM-Transformer model for enterprise workSmall · Open weights Zamba2 🇺🇸 United StatesZamba2 Hybrid Mamba2-transformer language model suite by ZyphraGeneral · Open weights · 2024 fnNemotron 3 Nano 🇺🇸 United StatesNemotron 3 Nano NVIDIA's 30B hybrid Mamba-Transformer MoE model for reasoning and agentsGeneral · Conditional licence · 2025 fnGranite 4.0 🇺🇸 United StatesGranite 4.0 IBM's hybrid Mamba/transformer language models for enterprise workGeneral · Open weights · 2025 Jamba 🇮🇱 IsraelJamba Hybrid open-weights language model with long contextGeneral · Open weights · 2024

The weekly digest

Every Monday: the model of the day, new pages on EveryLLM, and the conferences and anniversaries of the week ahead.