library

llama3.1

Llama 3.1 is a new state-of-the-art model from Meta available in 8B, 70B and 405B parameter sizes.

tools 8b 70b 405b

121.1M Pulls 93 Tags Updated 1 year ago

llama3.2

Meta's Llama 3.2 goes small with 1B and 3B models.

tools 1b 3b

87.6M Pulls 63 Tags Updated 2 years ago

mistral

The 7B model released by Mistral AI, updated to version 0.3.

tools 7b

34.2M Pulls 84 Tags Updated 1 year ago

llama3

Meta Llama 3: The most capable openly available LLM to date

8b 70b

26M Pulls 68 Tags Updated 2 years ago

mxbai-embed-large

State-of-the-art large embedding model from mixedbread.ai

embedding 335m

16.9M Pulls 4 Tags Updated 2 years ago

llava

🌋 LLaVA is a novel end-to-end trained large multimodal model that combines a vision encoder and Vicuna for general-purpose visual and language understanding. Updated to version 1.6.

vision 7b 13b 34b

15.4M Pulls 98 Tags Updated 2 years ago

glm-ocr

GLM-OCR is a multimodal OCR model for complex document understanding, built on the GLM-V encoder–decoder architecture.

vision tools

9.3M Pulls 3 Tags Updated 4 weeks ago

codellama

A large language model that can use text prompts to generate and discuss code.

7b 13b 34b 70b

8.2M Pulls 199 Tags Updated 2 years ago

llama2

Llama 2 is a collection of foundation language models ranging from 7B to 70B parameters.

7b 13b 70b

7.6M Pulls 102 Tags Updated 2 years ago

qwen3-vl

The most powerful vision-language model in the Qwen model family to date.

vision tools thinking 2b 4b 8b 30b 32b 235b

7.2M Pulls 57 Tags Updated 3 weeks ago

tinyllama

The TinyLlama project is an open endeavor to train a compact 1.1B Llama model on 3 trillion tokens.

1.1b

6.8M Pulls 36 Tags Updated 2 years ago

qwen2.5vl

Flagship vision-language model of Qwen and also a significant leap from the previous Qwen2-VL.

vision 3b 7b 32b 72b

6M Pulls 17 Tags Updated 1 year ago

mistral-nemo

A state-of-the-art 12B model with 128k context length, built by Mistral AI in collaboration with NVIDIA.

tools 12b

5.8M Pulls 17 Tags Updated 1 year ago

llama3.2-vision

Llama 3.2 Vision is a collection of instruction-tuned image reasoning generative models in 11B and 90B sizes.

vision 11b 90b

5.6M Pulls 9 Tags Updated 1 year ago

llama3.3

New state of the art 70B model. Llama 3.3 70B offers similar performance compared to the Llama 3.1 405B model.

tools 70b

4.5M Pulls 14 Tags Updated 1 year ago

dolphin3

Dolphin 3.0 Llama 3.1 8B 🐬 is the next generation of the Dolphin series of instruct-tuned models designed to be the ultimate general purpose local model, enabling coding, math, agentic, function calling, and general use cases.

8b

4.3M Pulls 5 Tags Updated 1 year ago

smollm2

SmolLM2 is a family of compact language models available in three size: 135M, 360M, and 1.7B parameters.

tools 135m 360m 1.7b

4.2M Pulls 49 Tags Updated 2 years ago

all-minilm

Embedding models on very large sentence level datasets.

embedding 22m 33m

3.9M Pulls 10 Tags Updated 2 years ago

olmo2

OLMo 2 is a new family of 7B and 13B models trained on up to 5T tokens. These models are on par with or better than equivalently sized fully open models, and competitive with open-weight models such as Llama 3.1 on English academic benchmarks.

7b 13b

3.8M Pulls 9 Tags Updated 1 year ago

snowflake-arctic-embed

A suite of text embedding models by Snowflake, optimized for performance.

embedding 22m 33m 110m 137m 335m

3.2M Pulls 16 Tags Updated 2 years ago

mistral-small

Mistral Small 3 sets a new benchmark in the “small” Large Language Models category below 70B.

tools 22b 24b

3.2M Pulls 21 Tags Updated 1 year ago

mixtral

A set of Mixture of Experts (MoE) model with open weights by Mistral AI in 8x7b and 8x22b parameter sizes.

tools 8x7b 8x22b

3M Pulls 70 Tags Updated 1 year ago

translategemma

A new collection of open translation models built on Gemma 3, helping people communicate across 55 languages.

vision 4b 12b 27b

2.9M Pulls 13 Tags Updated 10 months ago

llama2-uncensored

Uncensored Llama 2 model by George Sung and Jarrad Hope.

7b 70b

2.8M Pulls 34 Tags Updated 3 years ago

devstral-small-2

24B model that excels at using tools to explore codebases, editing multiple files and power software engineering agents.

vision tools 24b

2.7M Pulls 6 Tags Updated 11 months ago

falcon3

A family of efficient AI models under 10B parameters performant in science, math, and coding through innovative training techniques.

1b 3b 7b 10b

2.7M Pulls 17 Tags Updated 1 year ago

mistral-small3.2

An update to Mistral Small that improves on function calling, instruction following, and less repetition errors.

vision tools 24b

2.5M Pulls 5 Tags Updated 1 year ago

smollm

🪐 A family of small models with 135M, 360M, and 1.7B parameters, trained on a new high-quality dataset.

135m 360m 1.7b

2.4M Pulls 94 Tags Updated 2 years ago

sqlcoder

SQLCoder is a code completion model fined-tuned on StarCoder for SQL generation tasks

7b 15b

2.4M Pulls 48 Tags Updated 2 years ago

llava-llama3

A LLaVA model fine-tuned from Llama 3 Instruct with better scores in several benchmarks.

vision 8b

2.3M Pulls 4 Tags Updated 2 years ago

dolphin-llama3

Dolphin 2.9 is a new model with 8B and 70B sizes by Eric Hartford based on Llama 3 that has a variety of instruction, conversational, and coding skills.

8b 70b

2.2M Pulls 53 Tags Updated 2 years ago

glm-4.7-flash

As the strongest model in the 30B class, GLM-4.7-Flash offers a new option for lightweight deployment that balances performance and efficiency.

tools thinking

2.2M Pulls 4 Tags Updated 5 months ago

dolphin-mixtral

Uncensored, 8x7b and 8x22b fine-tuned models based on the Mixtral mixture of experts models that excels at coding tasks. Created by Eric Hartford.

8x7b 8x22b

2M Pulls 70 Tags Updated 1 year ago

llama4

Meta's latest collection of multimodal models.

vision tools 16x17b 128x17b

1.9M Pulls 11 Tags Updated 1 year ago

dolphin-mistral

The uncensored Dolphin model based on Mistral that excels at coding tasks. Updated to version 2.8.

7b

1.9M Pulls 120 Tags Updated 2 years ago

dolphin-phi

2.7B uncensored Dolphin model by Eric Hartford, based on the Phi language model by Microsoft Research.

2.7b

1.6M Pulls 15 Tags Updated 2 years ago

ministral-3

The Ministral 3 family is designed for edge deployment, capable of running on a wide range of hardware.

vision tools 3b 8b 14b

1.6M Pulls 13 Tags Updated 11 months ago

magistral

Magistral is a small, efficient reasoning model with 24B parameters.

tools thinking 24b

1.5M Pulls 5 Tags Updated 1 year ago

mistral-openorca

Mistral OpenOrca is a 7 billion parameter model, fine-tuned on top of the Mistral 7B model using the OpenOrca dataset.

7b

1.4M Pulls 17 Tags Updated 3 years ago

codestral

Codestral is Mistral AI’s first-ever code model designed for code generation tasks.

22b

1.4M Pulls 17 Tags Updated 2 years ago

mistral-large

Mistral Large 2 is Mistral's new flagship model that is significantly more capable in code generation, mathematics, and reasoning with 128k context window and support for dozens of languages.

tools 123b

1.4M Pulls 32 Tags Updated 1 year ago

deepseek-llm

An advanced language model crafted with 2 trillion bilingual tokens.

7b 67b

1.3M Pulls 64 Tags Updated 2 years ago

lfm2.5-thinking

LFM2.5 is a new family of hybrid models designed for on-device deployment.

tools thinking 1.2b

1.3M Pulls 5 Tags Updated 9 months ago

glm4

A strong multi-lingual general language model with competitive performance to Llama 3.

9b

1.3M Pulls 32 Tags Updated 2 years ago

deepscaler

A fine-tuned version of Deepseek-R1-Distilled-Qwen-1.5B that surpasses the performance of OpenAI’s o1-preview with just 1.5B parameters on popular math evaluations.

1.5b

1.3M Pulls 5 Tags Updated 1 year ago

neural-chat

A fine-tuned model based on Mistral with good coverage of domain and language.

7b

1.2M Pulls 50 Tags Updated 2 years ago

wizardlm2

State of the art large language model from Microsoft AI with improved performance on complex chat, multilingual, reasoning and agent use cases.

7b 8x22b

1.2M Pulls 22 Tags Updated 2 years ago

falcon

A large language model built by the Technology Innovation Institute (TII) for use in summarization, text generation, and chat bots.

7b 40b 180b

1.2M Pulls 38 Tags Updated 3 years ago

llama2-chinese

Llama 2 based model fine tuned to improve Chinese dialogue ability.

7b 13b

1.2M Pulls 35 Tags Updated 3 years ago

stablelm2

Stable LM 2 is a state-of-the-art 1.6B and 12B parameter language model trained on multilingual data in English, Spanish, German, Italian, French, Portuguese, and Dutch.

1.6b 12b

1.2M Pulls 84 Tags Updated 2 years ago

lfm2

LFM2 is a family of hybrid models designed for on-device deployment. LFM2-24B-A2B is the largest model in the family, scaling the architecture to 24 billion parameters while keeping inference efficient.

tools 24b

1.2M Pulls 6 Tags Updated 8 months ago

llava-phi3

A new small LLaVA model fine-tuned from Phi 3 Mini.

vision 3.8b

1.1M Pulls 4 Tags Updated 2 years ago

stable-code

Stable Code 3B is a coding model with instruct and code completion variants on par with models such as Code Llama 7B that are 2.5x larger.

3b

1.1M Pulls 36 Tags Updated 2 years ago

devstral

Devstral: the best open source model for coding agents

tools 24b

1.1M Pulls 5 Tags Updated 1 year ago

internlm2

InternLM2.5 is a 7B parameter model tailored for practical scenarios with outstanding reasoning capability.

1m 1.8b 7b 20b

1.1M Pulls 65 Tags Updated 2 years ago

llama-guard3

Llama Guard 3 is a series of models fine-tuned for content safety classification of LLM inputs and responses.

1b 8b

1.1M Pulls 33 Tags Updated 2 years ago

xwinlm

Conversational model based on Llama 2 that performs competitively on various benchmarks.

7b 13b

1.1M Pulls 80 Tags Updated 3 years ago

llama3-chatqa

A model from NVIDIA based on Llama 3 that excels at conversational question answering (QA) and retrieval-augmented generation (RAG).

8b 70b

1.1M Pulls 35 Tags Updated 2 years ago

dolphincoder

A 7B and 15B uncensored variant of the Dolphin model family that excels at coding, based on StarCoder2.

7b 15b

1.1M Pulls 35 Tags Updated 2 years ago

samantha-mistral

A companion assistant trained in philosophy, psychology, and personal relationships. Based on Mistral.

7b

1.1M Pulls 49 Tags Updated 3 years ago

yarn-llama2

An extension of Llama 2 that supports a context of up to 128k tokens.

7b 13b

1.1M Pulls 67 Tags Updated 3 years ago

llama3-gradient

This model extends LLama-3 8B's context length from 8k to over 1m tokens.

8b 70b

1.1M Pulls 35 Tags Updated 2 years ago

llama3-groq-tool-use

A series of models from Groq that represent a significant advancement in open-source AI capabilities for tool use/function calling.

tools 8b 70b

1.1M Pulls 33 Tags Updated 2 years ago

phind-codellama

Code generation model based on Code Llama.

34b

1M Pulls 49 Tags Updated 2 years ago

starling-lm

Starling is a large language model trained by reinforcement learning from AI feedback focused on improving chatbot helpfulness.

7b

1M Pulls 36 Tags Updated 2 years ago

solar

A compact, yet powerful 10.7B large language model designed for single-turn conversation.

10.7b

1M Pulls 32 Tags Updated 2 years ago

stable-beluga

Llama 2 based model fine tuned on an Orca-style dataset. Originally called Free Willy.

7b 13b 70b

1M Pulls 49 Tags Updated 3 years ago

shieldgemma

ShieldGemma is set of instruction tuned models for evaluating the safety of text prompt input and text output responses against a set of defined safety policies.

2b 9b 27b

1M Pulls 49 Tags Updated 2 years ago

wizardlm

General use model based on Llama 2.

986.8K Pulls 73 Tags Updated 3 years ago

reader-lm

A series of models that convert HTML content to Markdown content, which is useful for content conversion tasks.

0.5b 1.5b

979.5K Pulls 33 Tags Updated 2 years ago

paraphrase-multilingual

Sentence-transformers model that can be used for tasks like clustering or semantic search.

embedding 278m

966.5K Pulls 3 Tags Updated 2 years ago

llama-pro

An expansion of Llama 2 that specializes in integrating both general language understanding and domain-specific knowledge, particularly in programming and mathematics.

956.4K Pulls 33 Tags Updated 2 years ago

yarn-mistral

An extension of Mistral to support context windows of 64K or 128K.

7b

945.8K Pulls 33 Tags Updated 3 years ago

bakllava

BakLLaVA is a multimodal model consisting of the Mistral 7B base model augmented with the LLaVA architecture.

vision 7b

923.2K Pulls 17 Tags Updated 2 years ago

mistral-small3.1

Building upon Mistral Small 3, Mistral Small 3.1 (2503) adds state-of-the-art vision understanding and enhances long context capabilities up to 128k tokens without compromising text performance.

vision tools 24b

901.8K Pulls 5 Tags Updated 1 year ago

nemotron-3.5-lightning

NVIDIA Nemotron 3.5 Lightning is an open 30B mixture-of-experts (MoE) model with 3B active parameters built for always-on agents.

tools thinking 30b

886.3K Pulls 11 Tags Updated 3 weeks ago

command-r-plus

Command R+ is a powerful, scalable large language model purpose-built to excel at real-world enterprise use cases.

tools 104b

843.5K Pulls 21 Tags Updated 2 years ago

tinydolphin

An experimental 1.1B parameter model trained on the new Dolphin 2.8 dataset by Eric Hartford and based on TinyLlama.

1.1b

757.5K Pulls 18 Tags Updated 2 years ago

medllama2

Fine-tuned Llama 2 model to answer medical questions based on an open source medical dataset.

7b

754.9K Pulls 17 Tags Updated 3 years ago

wizardlm-uncensored

Uncensored version of Wizard LM model

13b

675.2K Pulls 18 Tags Updated 3 years ago

reflection

A high-performing model trained with a new technique called Reflection-tuning that teaches a LLM to detect mistakes in its reasoning and correct course.

70b

641.7K Pulls 17 Tags Updated 2 years ago

nous-hermes2-mixtral

The Nous Hermes 2 model from Nous Research, now trained over Mixtral.

8x7b

617.3K Pulls 18 Tags Updated 1 year ago

megadolphin

MegaDolphin-2.2-120b is a transformation of Dolphin-2.2-70b created by interleaving the model with itself.

120b

595.8K Pulls 19 Tags Updated 2 years ago

everythinglm

Uncensored Llama2 based model with support for a 16K context window.

13b

595.5K Pulls 18 Tags Updated 2 years ago

solar-pro

Solar Pro Preview: an advanced large language model (LLM) with 22 billion parameters designed to fit into a single GPU

22b

586.5K Pulls 18 Tags Updated 2 years ago

mathstral

MathΣtral: a 7B model designed for math reasoning and scientific discovery by Mistral AI.

7b

583.5K Pulls 17 Tags Updated 2 years ago

falcon2

Falcon2 is an 11B parameters causal decoder-only model built by TII and trained over 5T tokens.

11b

566.3K Pulls 17 Tags Updated 2 years ago

stablelm-zephyr

A lightweight chat model allowing accurate, and responsive output without requiring high-end hardware.

3b

562.8K Pulls 17 Tags Updated 2 years ago

duckdb-nsql

7B parameter text-to-SQL model made by MotherDuck and Numbers Station.

7b

554K Pulls 17 Tags Updated 2 years ago

mistrallite

MistralLite is a fine-tuned model based on Mistral with enhanced capabilities of processing long contexts.

7b

549.5K Pulls 17 Tags Updated 3 years ago

open-orca-platypus2

Merge of the Open Orca OpenChat model and the Garage-bAInd Platypus 2 model. Designed for chat and code generation.

13b

535.2K Pulls 17 Tags Updated 3 years ago

goliath

A language model created by combining two fine-tuned Llama 2 70B models into one.

506.4K Pulls 16 Tags Updated 2 years ago

snowflake-arctic-embed2

Snowflake's frontier embedding model. Arctic Embed 2.0 adds multilingual support without sacrificing English performance or scalability.

embedding 568m

500.5K Pulls 3 Tags Updated 1 year ago

olmo-3

Olmo is a series of Open language models designed to enable the science of language models. These models are pre-trained on the Dolma 3 dataset and post-trained on the Dolci datasets.

tools thinking 7b 32b

487.2K Pulls 15 Tags Updated 11 months ago

sailor2

Sailor2 are multilingual language models made for South-East Asia. Available in 1B, 8B, and 20B parameter sizes.

1b 8b 20b

443K Pulls 13 Tags Updated 1 year ago

mistral-medium-3.5

Mistral Medium 3.5 is the first flagship model of Mistral AI that merged instruction-following, reasoning, and coding in a single set of 128B weights.

vision tools thinking 128b

442.2K Pulls 5 Tags Updated 6 months ago

tulu3

Tülu 3 is a leading instruction following model family, offering fully open-source data, code, and recipes by the The Allen Institute for AI.

8b 70b

429.2K Pulls 9 Tags Updated 1 year ago

muse-glimmer

Meta's latest open model built for always-on local agents. 30B parameters, licensed under Apache 2.0 and runs on a single GPU — tuned for tool use, long tasks, and failure recovery.

vision tools thinking 30b

386.2K Pulls 15 Tags Updated 2 months ago

glm-5.2

GLM-5.2 is Z.ai’s flagship model for the era of long-horizon tasks.

tools thinking cloud

385.7K Pulls 1 Tag Updated 4 months ago

devstral-2

123B model that excels at using tools to explore codebases, editing multiple files and power software engineering agents.

tools 123b

380.9K Pulls 5 Tags Updated 11 months ago

alfred

A robust conversational model designed to be used for both chat and instruct use cases.

40b

341.4K Pulls 7 Tags Updated 2 years ago

lfm2.5

LFM2.5-8B-A1B, an edge model built for fast, reliable tool calling on consumer hardware.

tools thinking 8b 24b

334.5K Pulls 10 Tags Updated 4 weeks ago

olmo-3.1

Olmo is a series of Open language models designed to enable the science of language models. These models are pre-trained on the Dolma 3 dataset and post-trained on the Dolci datasets.

tools thinking 32b

320.7K Pulls 10 Tags Updated 11 months ago

bge-large

Embedding model from BAAI mapping texts to vectors.

embedding 335m

306.6K Pulls 3 Tags Updated 2 years ago

smallthinker

A new small reasoning model fine-tuned from the Qwen 2.5 3B Instruct model.

3b

276.6K Pulls 5 Tags Updated 1 year ago

qwen3.8-flash-next

This experimental preview of the architecture that will underpin Qwen4.

vision tools thinking

252.2K Pulls 6 Tags Updated 3 days ago

glm-5.3

Z.ai's flagship model and the most capable open-weights model for coding, with major gains on long-horizon agentic tasks.

tools thinking cloud

248.1K Pulls 1 Tag Updated 2 months ago

deepseek-v4.1-flash

Designed for greater capability, faster inference, higher throughput, and scaling to larger models.

vision tools thinking cloud

228.1K Pulls 2 Tags Updated 3 weeks ago

glm-5.3-flash

Z.ai's first natively multimodal model, approaching Claude Opus 4.8 on coding and agentic benchmarks with just 18B active parameters.

vision tools thinking cloud

219.1K Pulls 1 Tag Updated 2 months ago

laguna-s-2.1

Our most capable model to date, designed for long-horizon work. 70.2% on Terminal-Bench 2.1 at 118B-A8B.

tools thinking

194.4K Pulls 8 Tags Updated 2 months ago

laguna-xs-2.1

Laguna XS 2.1 is a 33B total parameter Mixture-of-Experts model with 3B activated parameters per token designed for agentic coding and long-horizon work on a local machine.

tools thinking

154.5K Pulls 7 Tags Updated 2 months ago

nemotron-3-ultra

NVIDIA Nemotron 3 Ultra is built for high-throughput reasoning and long-running agent workflows.

tools thinking cloud

154.4K Pulls 1 Tag Updated 5 months ago

clef

Clef is a 27B multimodal model created by Cloudflare that turns a state and a schema of typed questions into decisions.

vision decision cloud 27b

100.2K Pulls 9 Tags Updated 1 month ago

nimble

A 9B decision model from Bespoke Labs for fast, typed classification.

tools decision thinking 9b

93.4K Pulls 8 Tags Updated 1 month ago

flint

Flint is a 39.8 billion-parameter mixture-of-experts model from Recursive, built for coding agents.

tools thinking 40b

86.5K Pulls 7 Tags Updated 3 weeks ago

clef-flash

Clef-Flash is a 9B multimodal model that turns a state and a schema of typed questions into decisions.

vision decision 9b

79.7K Pulls 5 Tags Updated 1 month ago

lfm3

LFM3-1.7B is a tri-hybrid model for local agents and specialized applications.

tools thinking 1.7b

79.7K Pulls 9 Tags Updated 2 weeks ago

nemotron-3.5-bolt

Nemotron 3.5 Bolt is NVIDIA’s compact vision-language model for local coding, reasoning, and tool use.

vision tools thinking audio e4b

65.9K Pulls 10 Tags Updated 2 weeks ago

lfm3-omni

a single end-to-end model that understands text, images, video and audio, and answers in text and speech

vision tools audio 2b

60.2K Pulls 8 Tags Updated 2 weeks ago

minimax-m3.1-flash

As a next-generation Flash model, M3.1 Flash delivers a generational leap in agentic tasks, coding, Computer Use and cybersecurity, while making online inference nearly three times faster and reducing usage costs by 70%.

vision tools thinking cloud

53K Pulls 1 Tag Updated 3 weeks ago

laguna-xs.2

Laguna XS.2 is a 33B total parameter Mixture-of-Experts model with 3B activated parameters per token designed for agentic coding and long-horizon work on a local machine.

tools thinking

45.6K Pulls 7 Tags Updated 3 months ago

lfm3-vl

LFM3-VL-2B is a vision-language model for on-device visual understanding.

vision tools thinking 2b

43.7K Pulls 9 Tags Updated 2 weeks ago

pplx-decider

pplx-decider is a 27B decision model from Perplexity, fine-tuned from Qwen3.8-27B

vision decision 27b

39.2K Pulls 6 Tags Updated 3 weeks ago

lfm3-audio

LFM3-Audio-2B is a compact, fully featured audio model.

vision tools audio 2b

37.4K Pulls 8 Tags Updated 2 weeks ago

mistral-large-4

Mistral Large 4 is Mistral's open-weight, general-purpose multimodal model. Its granular Mixture-of-Experts design activates 52B of 1.05T total parameters, with a 1.6B vision encoder handling images.

vision tools thinking cloud

30.8K Pulls 1 Tag Updated 1 month ago