Models For Light And Fast Inference Models that are nice for extremely bulk text or slow devices. prism-ml/Bonsai-1.7B-gguf Text Generation • 2B • Updated Apr 18 • 153k • 97 LiquidAI/LFM2.5-230M Text Generation • 0.2B • Updated Aug 19 • 31.1k • 321 google/gemma-3-270m-it Text Generation • 0.3B • Updated Aug 14, 2025 • 58.2k • 712 HuggingFaceTB/SmolLM2-135M-Instruct Text Generation • 0.1B • Updated Sep 22, 2025 • 848k • 471
Best 2B-Class Models 1B just sucks, 2B is bare minimum for day-to-day use, 4B is way better if possible google/gemma-4-E2B-it Any-to-Any • 5B • Updated 5 days ago • 3.17M • 1.12k LiquidAI/LFM2.5-2.6B Text Generation • 3B • Updated 6 days ago • 52k • 859 Qwen/Qwen3.5-2B-Base Image-Text-to-Text • 2B • Updated Apr 23 • 329k • 108 openbmb/MiniCPM5-2B Text Generation • 3B • Updated Sep 29 • 1.06M • 1.84k
Models For Light And Fast Inference Models that are nice for extremely bulk text or slow devices. prism-ml/Bonsai-1.7B-gguf Text Generation • 2B • Updated Apr 18 • 153k • 97 LiquidAI/LFM2.5-230M Text Generation • 0.2B • Updated Aug 19 • 31.1k • 321 google/gemma-3-270m-it Text Generation • 0.3B • Updated Aug 14, 2025 • 58.2k • 712 HuggingFaceTB/SmolLM2-135M-Instruct Text Generation • 0.1B • Updated Sep 22, 2025 • 848k • 471
Best 2B-Class Models 1B just sucks, 2B is bare minimum for day-to-day use, 4B is way better if possible google/gemma-4-E2B-it Any-to-Any • 5B • Updated 5 days ago • 3.17M • 1.12k LiquidAI/LFM2.5-2.6B Text Generation • 3B • Updated 6 days ago • 52k • 859 Qwen/Qwen3.5-2B-Base Image-Text-to-Text • 2B • Updated Apr 23 • 329k • 108 openbmb/MiniCPM5-2B Text Generation • 3B • Updated Sep 29 • 1.06M • 1.84k