LLM & Models Tools
🤖
Yi Models
Freemium
Yi is a family of large language models from 01.AI founded by Kai-Fu Lee, offering strong bilingual Chinese-English performance. Yi-34B and Yi-large achieve competitive performance with frontier models on reasoning and knowledge benchmarks. Yi models are available as open-source weights for research and commercial deployment. The Yi-VL vision-language model adds image understanding capabilities. Used by researchers and enterprises in Asia and globally for applications requiring strong Chinese language capabilities alongside English proficiency.
🤖
Falcon
Free
Falcon is a family of open-source large language models developed by the Technology Innovation Institute (TII) in Abu Dhabi. Falcon 180B was the world's largest open-source model at launch and achieves performance competitive with proprietary models. Falcon models are trained on the RefinedWeb dataset and released under permissive licences for commercial use. Falcon 2 adds multimodal vision capabilities. Used by researchers and enterprises requiring powerful open-source models with Middle Eastern institutional backing and no usage restrictions.
🤖
Vicuna
Free
Vicuna is an open-source chatbot fine-tuned from Llama using user-shared conversations from ShareGPT. Developed by researchers at UC Berkeley, CMU, Stanford, and UCSD, Vicuna 13B achieved 90% of ChatGPT quality in early evaluations at a fraction of the cost. It pioneered the approach of fine-tuning open-source base models on human conversation data to create capable chat assistants. Vicuna's methodology influenced subsequent open-source chat model development and remains an important reference in the open LLM ecosystem.
🤖
SOLAR
Freemium
SOLAR 10.7B is a compact yet powerful language model from Upstage AI that achieves performance surpassing Llama 2 70B and Mistral 7B despite its smaller size through depth up-scaling, a novel model construction technique. SOLAR is commercially available and achieves state-of-the-art performance among sub-11B parameter models. Upstage's Document AI products are built on SOLAR technology. Used by developers and enterprises needing high-performance language models that can run on modest hardware without sacrificing quality compared to much larger models.
🤖
Nous Hermes
Free
Nous Hermes is a series of fine-tuned language models by Nous Research that consistently rank among the top performers on open LLM leaderboards. Built on Llama and Mistral base models, Hermes models are fine-tuned on high-quality instruction datasets emphasising reasoning, code generation, and structured output following. Nous Hermes 2 Pro adds function calling and structured JSON output capabilities. Popular in the open-source AI community for applications requiring reliable instruction following and reasoning without proprietary model dependencies.
🤖
OpenHermes
Free
OpenHermes is a series of high-quality fine-tuned language models from Teknium and the open-source AI community built on Mistral and Llama base models. OpenHermes 2.5 Mistral 7B is one of the most downloaded models on HuggingFace, achieving remarkable instruction following, coding, and reasoning performance for its size. Trained entirely on synthetic and curated open-source datasets, OpenHermes demonstrates that high-quality training data matters more than model scale for many tasks. Widely used for local AI applications and edge deployment.
🤖
StableLM
Free
StableLM is Stability AI's family of open-source language models ranging from 1.6B to 12B parameters designed for efficient local deployment. StableLM 2 12B achieves strong performance relative to its size and runs well on consumer hardware. The models are trained on diverse, high-quality datasets and released under permissive licences. Stability AI's focus on smaller, efficient models makes StableLM particularly useful for on-device applications, edge AI deployment, and scenarios where inference cost and latency are primary constraints.
🤖
Code Llama
Free
Code Llama is Meta's specialised code generation model fine-tuned from Llama 2 on code-heavy datasets. Available in 7B, 13B, 34B, and 70B sizes with specialised variants for Python and instruction-following. Code Llama achieves state-of-the-art performance on HumanEval and MBPP coding benchmarks among open-source models. It supports infilling for code completion within existing code. Used by developers building local code assistants, IDE plugins, and coding tools who want powerful code generation without API costs or code privacy concerns.
🤖
WizardLM
Free
WizardLM is a series of instruction-tuned language models from Microsoft Research using Evol-Instruct, a novel method of automatically evolving simple instructions into complex ones for training data generation. WizardLM 2 8x22B achieves performance competitive with GPT-4 on instruction following benchmarks. WizardCoder specialises in code generation and WizardMath in mathematical reasoning. The Evol-Instruct methodology has influenced subsequent open-source model training approaches for creating complex, diverse instruction datasets.
🤖
Orca 2
Free
Orca 2 is a small language model from Microsoft Research that achieves performance far exceeding its size through careful reasoning capability training. Using a technique called explanation tuning where the model learns different reasoning strategies for different tasks, Orca 2 13B outperforms much larger models on complex reasoning benchmarks. Microsoft Research designed Orca 2 to demonstrate that smaller models can match large model reasoning when trained appropriately. Published with open weights for research use.
🤖
InternLM
Free
InternLM is a series of open-source language models from Shanghai AI Laboratory achieving strong performance across English and Chinese tasks. InternLM2 20B achieves competitive performance with much larger models through careful training on diverse, high-quality datasets. The InternLM series includes specialised variants for mathematics, coding, and tool use. Shanghai AI Lab's open research approach has made InternLM popular among Chinese AI researchers and enterprises requiring strong bilingual capabilities with full model access for fine-tuning.
🤖
Baichuan
Free
Baichuan is a series of large language models from Baichuan Intelligence in China, trained on both Chinese and English data with strong performance across knowledge-intensive tasks. Baichuan 2 achieves top results among Chinese open-source models on Chinese language benchmarks. The models are available in 7B and 13B sizes with chat-tuned variants for conversational applications. Used by Chinese enterprises and researchers requiring high-quality bilingual AI models with Chinese cultural and linguistic knowledge not well represented in Western training datasets.
Browse Other Categories
Image Generation
Video AI
Productivity
AI Tool
Writing & Content
Audio & Music
Code & Developer
AI Companion
Gaming AI
Data & Analytics
Finance
Framework
Marketing
Education
Legal
MLOps
Security
Directory
E-commerce
AI Agents
APIs
Automation
Cybersecurity AI
Database
Healthcare AI
HR & Recruiting
NLP
Platform
Real Estate AI
Research
Search