Open-source community models on Hugging Face
| Model | API ID | Release | Cutoff | Conf. | Status | Notes |
|---|---|---|---|---|---|---|
| GPT-OSS (20B/120B) | openai/gpt-oss-20b | 2025-08 | 2025-06 | medium | active | OpenAI's first open-weight model. Apache 2.0. Also on Groq. |
| SmolLM3 (3B) | HuggingFaceTB/SmolLM3-3B | 2025-07 | 2025-04 | low | active | HuggingFace TB. |
| BLOOM (176B) | bigscience/bloom | 2022-07 | 2022-03 | low | active | BigScience. 46 natural + 13 programming languages. |
| InternLM2 (7B/20B) | internlm/internlm2-chat-7b | 2024-01 | 2023-09 | low | active | Shanghai AI Lab. |
| InternVL2 (8B-108B) | OpenGVLab/InternVL2-8B | 2024-07 | 2024-04 | low | active | Vision-language model series. |
| MPT-7B | mosaicml/mpt-7b-instruct | 2023-05 | 2023-02 | low | active | MosaicML. Commercial license. |
| MPT-30B | mosaicml/mpt-30b-instruct | 2023-06 | 2023-03 | low | active | |
| Pythia (1B-12B) | EleutherAI/pythia-12b | 2023-01 | 2022-12 | low | active | EleutherAI research model. |
| GPT-NeoX (20B) | EleutherAI/gpt-neox-20b | 2022-04 | 2022-01 | low | active | EleutherAI. |
| BLOOMZ | bigscience/bloomz | 2022-11 | 2022-07 | low | active | Instruction-tuned BLOOM. |
| InternVL3 (1B-78B) | OpenGVLab/InternVL3-8B | 2025-03 | 2025-01 | low | active | Latest InternVL. |
| SmolLM2 (135M-1.7B) | HuggingFaceTB/SmolLM2-1.7B-Instruct | 2024-11 | 2024-08 | low | active | HuggingFace TB. Ultra-compact. |
| Gemma 3n (2B/4B) | google/gemma-3n-E4B-it | 2025-06 | 2025-03 | low | active | On-device Gemma variant. |
| DeepSeek-VL2 | deepseek-ai/deepseek-vl2 | 2024-12 | 2024-07 | low | active | MoE vision-language model. |
| Qwen2.5-VL (3B/7B/72B) | Qwen/Qwen2.5-VL-7B-Instruct | 2025-01 | 2024-10 | low | active | Vision-language. |
| Pixtral Large (124B) | mistralai/Pixtral-Large-Instruct-2411 | 2024-11 | 2024-07 | low | active | Mistral's large multimodal. |
| Kimi-VL (3B/16B MoE) | moonshotai/Kimi-VL-A3B-Instruct | 2025-04 | 2025-01 | low | active | Moonshot vision-language. |
| Danube2 (1.8B) | h2oai/h2o-danube2-1-8b-chat | 2024-01 | 2023-10 | low | active | H2O.ai small model. |
| TIGER-Lab Mantis (8B) | TIGER-Lab/Mantis-8B-siglip-llama3 | 2024-06 | 2024-03 | low | active | Multi-image multimodal. |
Knowledge Cutoffs