Submit Tool
#

Open-Weight

AI tools tagged Open-Weight.

200+ tools · handpicked & curated
// price
ALBERT Google's parameter-sharing variant of BERT designed for efficiency at scale
Aquila BAAI's first open bilingual (Chinese/English) foundation model
Aquila2 BAAI's second-generation bilingual foundation model family
Arctic Snowflake's open-weight enterprise-focused model, optimized for SQL and RAG
Arctic Embed 2.0 Snowflake's open-weight multilingual retrieval embedding model
AudioCraft Meta's open-source framework spanning music and sound-effect generation
Aya 23 Cohere For AI's open-weight multilingual model covering 23 languages
Baichuan 2 Baichuan's open-weight bilingual foundation model
Baichuan-13B Baichuan's open-weight bilingual model, predating the Baichuan 2/3/4 generations
BERT Google's 2018 bidirectional encoder that reshaped NLP research
BGE-large-en BAAI's widely-used general embedding model for English retrieval
BGE-M3 BAAI's open-weight multilingual, multi-granularity embedding model
BioGPT Microsoft's domain-specific generative model pre-trained on biomedical literature
BitNet b1.58 Microsoft's research model demonstrating competitive LLM performance with 1.58-bit ternary weights
BLIP Salesforce's vision-language pretraining model for captioning and retrieval
BLIP-2 Salesforce's more efficient successor to BLIP, bootstrapping vision-language pretraining from frozen encoders
BLOOM The open, multilingual model built by the BigScience collaboration
BLOOMZ An instruction-tuned variant of BLOOM trained to follow cross-lingual instructions
BTLM-3B Cerebras' compact open-weight model tuned to punch above its parameter count
ByT5 Google's token-free variant of T5 that operates directly on raw bytes
Canary NVIDIA's multilingual speech recognition and translation model
Cerebras-GPT Cerebras' fully open family of models trained on its wafer-scale chips
Chameleon Meta's early-fusion model that natively mixes text and image tokens
ChatGLM3 Zhipu's third-generation open-weight bilingual dialogue model
CLIP OpenAI's foundational image-text contrastive model that underlies much of modern multimodal AI
CLIP (LAION variant) LAION's open-weight retrained CLIP vision-language embedding model
Code Llama Meta's open-weight code-specialized Llama variant
CodeGen2 Salesforce's open-weight code model built for infilling and multi-turn program synthesis
Codestral Mistral's code-generation and completion model
Codestral Mamba Mistral's code model built on the Mamba state-space architecture instead of a transformer
CodeT5+ Salesforce's open-weight encoder-decoder code model family
CogVideoX Zhipu's open-weight text-to-video model family
CogView3 Zhipu's relay-diffusion text-to-image model
Command R Plus Cohere's flagship RAG- and tool-use-optimized model above Command R
Command R+ Cohere's retrieval-augmented-generation-optimized model
Command R7B Cohere's smallest current-generation Command model
CosyVoice Alibaba's open-weight zero-shot voice-cloning and text-to-speech model
DBRX Databricks' open-weight mixture-of-experts model
DBRX Instruct Instruction-tuned variant of Databricks' open-weight DBRX model
DeepFloyd IF A pixel-space (non-latent) text-to-image model developed by the Stability-backed DeepFloyd lab
DeepSeek Coder V2 DeepSeek's open-weight code-specialized mixture-of-experts model
DeepSeek LLM 67B DeepSeek's first dense open-weight foundation model
DeepSeek R1 DeepSeek's reasoning model that matched closed frontier models on math and code
DeepSeek V3 DeepSeek's efficient mixture-of-experts model trained at a fraction of rival costs
DeepSeek V4-Pro The top open-weight model of 2026, leading on agentic coding and graduate reasoning
DeepSeek-Coder-33B DeepSeek's earlier flagship code model, preceding DeepSeek-Coder-V2
DeepSeek-Math DeepSeek's open-weight model specialized for mathematical reasoning
DeepSeek-MoE-16B DeepSeek's early fine-grained mixture-of-experts model that informed its later MoE architectures
DeepSeek-R1-Distill-Llama-70B A Llama-based distillation of DeepSeek-R1's reasoning traces, the largest distilled variant
DeepSeek-R1-Distill-Qwen-32B A Qwen-based distillation of DeepSeek-R1's reasoning traces into a smaller open model
DeepSeek-V2 DeepSeek's 2024 mixture-of-experts model that undercut rivals on API price
DeepSeek-VL DeepSeek's first vision-language model, preceding DeepSeek-VL2
DeepSeek-VL2 DeepSeek's open-weight vision-language mixture-of-experts model
Devstral Mistral's open-weight model tuned for agentic coding and codebase exploration
DINOv2 Meta's self-supervised vision model producing general-purpose image features without labels
DistilBERT Hugging Face's distilled, 40%-smaller version of BERT that retains most of its performance
Dolly 2.0 Databricks' first fully open, commercially-usable instruction-following model
ELECTRA Google's more sample-efficient pretraining approach using a replaced-token-detection objective
ELMo AI2's deep contextualized word representation model that predated the transformer era
Emu3 BAAI's next-token-prediction model unifying text, image, and video generation and understanding
ERNIE 4.5 Baidu's flagship multimodal foundation model
ESM-2 Meta's protein language model used to predict structure from sequence
EXAONE 3.0 LG's earlier EXAONE generation, preceding EXAONE 3.5
EXAONE 3.5 LG's open-weight bilingual Korean-English model family
Falcon 180B TII's open-weight model that led open leaderboards on release
Falcon 2 TII's Falcon generation between Falcon 180B and Falcon 3, adding a vision-language variant
Falcon 3 TII's efficient small-model open-weight family
Falcon 40B TII's earlier open-weight model that led leaderboards before Falcon 180B
Falcon 7B TII's compact Falcon tier that helped popularize the original Falcon release
Falcon Mamba 7B TII's open-weight model built on the state-space Mamba architecture
Flan-T5 Google's instruction-tuned, open-weight successor to T5
Florence-2 Microsoft's compact vision foundation model for captioning, detection, and segmentation in one model
FLUX.1 dev Black Forest Labs' open-weight tier of FLUX for non-commercial use
FLUX.1 Kontext Black Forest Labs' image editing model that follows instructions on an existing image
FLUX.1 Krea Black Forest Labs and Krea's collaborative model tuned against over-stylised 'AI look'
Fuyu-8B Adept's open-weight multimodal model built for digital-agent perception
Galactica Meta's 2022 model trained on scientific literature, pulled days after launch
Gemma Google's original open-weight Gemma release, built from the same research as Gemini
Gemma 2 Google's prior-generation open-weight model family
Gemma 3 Google's open-weight model family built from Gemini research
GLM-4 Zhipu's 2024 general-purpose open-weight model
GLM-4.6 Zhipu's value-tier open-weight model for day-to-day coding
GLM-5.2 Zhipu's flagship open-weight model, shipped alongside Kimi K2.7 in mid-2026
GPT-2 OpenAI's 2019 model, once withheld from release over misuse concerns
GPT-J EleutherAI's early open-weight GPT-3-style model
GPT-Neo 2.7B EleutherAI's early GPT-3-style open model, a predecessor to GPT-J and GPT-NeoX
GPT-NeoX-20B EleutherAI's open-weight model, one of the largest public checkpoints of its era
GPT-OSS-120B OpenAI's first open-weight model release since GPT-2, matching o4-mini on many reasoning benchmarks
GPT-OSS-20B The smaller, single-GPU-friendly tier of OpenAI's open-weight model release
Granite 3.0 IBM's open-weight enterprise model family for regulated industries
Granite Code IBM's open-weight code generation model family
Granite Embedding IBM's enterprise text embedding model family for retrieval
Granite Guardian IBM's safety and risk-detection model for filtering harmful LLM inputs and outputs
Granite Vision IBM's compact vision-language model for document and chart understanding
Grok-1 xAI's first model, later open-weighted under Apache 2.0
GTE-large Alibaba's general text embedding model for retrieval and semantic similarity
Hermes 2 Nous Research's earlier open-weight fine-tune line
Hermes 3 Nous Research's open-weight fine-tune focused on steerability and reasoning
Hunyuan DiT Tencent's open-weight diffusion transformer for Chinese and English text-to-image
Hunyuan Large Tencent's open-weight mixture-of-experts model
Hunyuan Video Tencent's open-weight text-to-video generation model
Hunyuan-A13B Tencent's open-weight mixture-of-experts reasoning model
Hunyuan3D Tencent's open-weight model for generating 3D assets from text or images
IDEFICS2 Hugging Face's open vision-language model built for document and chart QA
InternLM2 Shanghai AI Lab's open-weight foundation model with strong long-context and reasoning ability
InternLM2.5 Shanghai AI Lab's updated InternLM generation with stronger tool-use and reasoning
InternVL2 Shanghai AI Lab's open vision-language model family, competitive with proprietary multimodal models
Jamba 1.5 Large AI21's hybrid transformer-Mamba model built for very long context
Jamba 1.5 Mini AI21's compact hybrid transformer-Mamba model
Jamba Large 1.6 AI21's updated flagship hybrid transformer-Mamba model
Jina Embeddings v2 Jina AI's embedding model, notable for an 8k-token context window at launch
jina-embeddings-v3 Jina AI's open-weight multilingual embedding model
Kimi K2 Moonshot's original K2-generation open-weight base model
Kimi K2.6 Moonshot's open-weight model built for long-running agentic and tool-use tasks
Kimi K2.7 Code Moonshot's open-weight agentic coding model, tuned for tool-use and long-running tasks
Kimi-Audio Moonshot AI's open-weight audio foundation model for understanding and generating speech
Kimi-VL Moonshot's open-weight vision-language model
Kolors Kuaishou's open-weight text-to-image model, strong on Chinese-language prompts
Kosmos-2 Microsoft's grounded multimodal model that links language to specific image regions
LLaMA Meta's original 2023 open-weight release that kicked off the open LLM boom
Llama 2 Meta's 2023 open-weight release that jump-started the open LLM ecosystem
Llama 3 Meta's 2024 open-weight release that closed most of the gap to closed models
Llama 3.1 405B Meta's largest dense open-weight model, competitive with closed frontier models
Llama 3.1 70B The mid-sized tier of Meta's Llama 3.1 release, between the 8B and 405B models
Llama 3.1 8B Meta's compact tier of the Llama 3.1 generation
Llama 3.2 Meta's first Llama generation with vision-capable and on-device tiers
Llama 3.3 Meta's efficient 70B open-weight model matching larger predecessors
Llama 4 Maverick Meta's flagship open-weight multimodal mixture-of-experts model
Llama 4 Scout Meta's smaller, faster Llama 4 tier with a 10M-token context window
Llama Guard 2 Meta's earlier content-safety classifier model, preceding Llama Guard 3
Llama Guard 3 Meta's open-weight safety classifier model for moderating LLM input and output
LLeMA-7B EleutherAI's open model continually pretrained on math and science text
Mathstral Mistral's math-specialised model derived from Mistral 7B
MiniCPM OpenBMB's efficient small model designed to run well on phones and edge devices
MiniCPM-V OpenBMB's multimodal edge model, capable of GPT-4V-level vision tasks on a phone
MiniMax M3 First open-weight model to combine frontier coding, a 1M-token context, and native multimodality
MiniMax-01 MiniMax's 2025 open-weight model with a 4M-token context window
Ministral 8B Mistral's compact model built for on-device and edge deployment
Mistral 7B Mistral's original compact open-weight model that outperformed larger rivals
Mistral NeMo Mistral's 12B open-weight model built with NVIDIA
Mistral Small 3 Mistral's efficient small model tier for latency-sensitive workloads
Mixtral 8x22B Mistral's open-weight sparse mixture-of-experts model
Mixtral 8x7B Mistral's original open-weight mixture-of-experts model that popularised sparse MoE LLMs
MPT-30B MosaicML's open-weight commercial-use model, later folded into Databricks
mT5 Google's multilingual variant of T5, covering 101 languages
MusicGen Meta's open-weight controllable music generation model
Nemotron Nano NVIDIA's small, efficient Nemotron tier optimised for edge deployment
Nemotron-3 8B NVIDIA's earlier compact enterprise-focused Nemotron model
Nemotron-4 340B NVIDIA's open-weight model built to generate synthetic training data
NLLB-200 Meta's No Language Left Behind model translating between 200 languages, including many low-resource ones
Nomic Embed Text v1 Nomic's first fully open-source, open-weight, open-data long-context embedding model
nomic-embed-text-v2 Nomic AI's fully open, reproducible text embedding model
NVIDIA Cosmos NVIDIA's world-foundation model family for physical AI and autonomous simulation
OLMo 2 AI2's fully open model, released with training data and code alongside weights
OLMo 7B AI2's first fully open model release, with open weights, data, and training code
OLMoE AI2's fully open mixture-of-experts model
OPT-175B Meta's 2022 open-weight model released to mirror GPT-3's scale
Orca 2 Microsoft Research's reasoning-focused fine-tune of Llama 2
Orca-Math Microsoft's small model specialised in grade-school math word problems
Pharia-1 Aleph Alpha's open-weight European-sovereignty-focused model
Phi-2 Microsoft's earlier 2.7B model demonstrating outsized small-model performance
Phi-3 Microsoft's compact open-weight model family for on-device use
Phi-3-medium The 14B-parameter tier of Microsoft's Phi-3 family, tuned for stronger reasoning
Phi-3-small The 7B-parameter tier of Microsoft's Phi-3 small-language-model family
Phi-3-vision Multimodal member of the Phi-3 family for image reasoning on-device
Phi-3.5-mini Microsoft's updated compact model in the Phi-3 line
Phi-4 Microsoft's small reasoning-focused model that punches above its parameter count
Phi-4-reasoning Reasoning-tuned variant of Phi-4 trained with chain-of-thought supervision
Pixtral 12B Mistral's first multimodal model, smaller than Pixtral Large
Pixtral Large Mistral's flagship vision-language model
Playground v2.5 Playground's aesthetics-tuned open-weight image model
ProGen Salesforce's generative model for designing novel protein sequences
Pythia EleutherAI's fully open suite of models built for interpretability research
Qwen-7B Alibaba's original Qwen release that started the model family
Qwen-Audio Alibaba's first audio-language model for speech and sound understanding
Qwen-Image Alibaba's text-to-image model built on the Qwen architecture, strong at rendering embedded text
Qwen-VL Alibaba's first vision-language model in the Qwen family
Qwen1.5-110B Alibaba's largest dense model in the Qwen1.5 generation
Qwen2-72B Alibaba's second-generation flagship dense open-weight model
Qwen2-Audio Alibaba's second-generation audio-language model with voice-chat capability
Qwen2.5-72B Alibaba's widely-deployed dense open-weight model
Qwen2.5-Coder Alibaba's dedicated code model line, competitive with much larger general models on coding benchmarks
Qwen2.5-VL Alibaba's open-weight vision-language model
Qwen3-235B-A22B Alibaba's flagship open-weight mixture-of-experts model, the cheapest capable frontier model
Qwen3-Coder Alibaba's open-weight agentic coding model
QwQ-32B Alibaba's open-weight reasoning model competitive with much larger models
QwQ-32B-Preview Alibaba's first public preview of its QwQ reasoning line, ahead of the full QwQ-32B release
RedPajama-INCITE Together AI's fully open model trained on the reproduced RedPajama dataset
Reka Edge Reka's smallest, on-device-friendly model tier
replit-code-v1.5 Replit's open-weight code-completion model
RoBERTa Meta's robustly-optimized retraining of BERT that improved on it across benchmarks
SAM Meta's Segment Anything Model, a foundation model for image segmentation
SAM 2 Meta's successor to SAM, extending segmentation to video
SantaCoder BigCode's small multilingual code model preceding StarCoder
Seamless M4T Meta's unified model for speech-to-speech and speech-to-text translation across ~100 languages
Seamless M4T v2 Improved release of Meta's massively multilingual speech translation model
Skywork-13B Kunlun Tech's open-weight bilingual foundation model
SmolLM Hugging Face's original family of small, fully open language models
SmolLM2 Hugging Face's fully open small-model family for on-device use
Solar 10.7B Upstage's depth-upscaled open model that preceded the Solar Pro line
Solar Pro 2 Upstage's efficient open-weight model that punches above its parameter count
Stable Cascade Stability AI's highly compressed-latent-space image model for faster training and inference
Stable Diffusion 3.5 Stability AI's open-weight text-to-image diffusion model
Stable Diffusion XL Stability AI's widely-adopted 2023 image generation model
Stable Video Diffusion Stability AI's first open-weight image-to-video model
StableLM 2 Stability AI's open-weight small language model
StableLM 3B Stability AI's early compact open-weight language model, preceding StableLM 2
StableLM Zephyr Stability AI's chat-tuned small model built on StableLM 3B
StarCoder The original BigCode model that StarCoder2 later succeeded
StarCoder2 The BigCode collaboration's open-weight code generation model
T5 Google's text-to-text transfer transformer, an early unifying NLP framework
Tacotron 2 Google's neural text-to-speech model pairing a spectrogram predictor with a WaveNet vocoder
UL2 Google's unified pretraining framework blending multiple denoising objectives
Vicuna-13B LMSYS's influential early open chat model fine-tuned from Llama on ShareGPT conversations
Vicuna-33B The largest Vicuna release, LMSYS's ShareGPT fine-tune of Llama
Wan2.1 Alibaba's open-weight video generation model family
Wav2Vec 2.0 Meta's self-supervised speech representation model that underpins many ASR systems
Whisper OpenAI's open-weight speech recognition and transcription model
Whisper Large v3 The largest, most accurate tier of OpenAI's open-weight speech recognition model
WizardCoder Microsoft's code-specialised fine-tune built with Evol-Instruct data generation
WizardLM-2 Microsoft's instruction-tuned model family built with complex, evolved synthetic training data
XGen-7B Salesforce's open-weight long-context research model
XLNet A permutation-based autoregressive pretraining model that outperformed BERT on many tasks
Yi-1.5-34B 01.AI's improved open-weight Yi generation
Yi-34B 01.AI's open-weight bilingual model
Yi-Coder 01.AI's open-weight code-specialized model
Yi-VL-34B 01.AI's open-weight vision-language model built on Yi-34B
Yuan 2.0 Inspur's open-weight foundation model tuned for enterprise Chinese-language tasks
Zephyr 141B Hugging Face's larger mixture-of-experts Zephyr fine-tune, distinct from the 7B release
Zephyr 7B Hugging Face's open-weight fine-tune that popularized DPO alignment