LLM & Models Tools
🤖
XVERSE
Free
XVERSE is a multilingual large language model series from XVERSE Technology in China supporting 40+ languages with particularly strong Chinese and English performance. XVERSE-65B achieves competitive performance with international frontier models on Chinese language benchmarks. The model series covers dense and mixture-of-experts architectures across multiple sizes. Used by enterprises in China and Southeast Asia requiring AI models with strong multilingual capabilities spanning Asian languages not well served by Western-developed models.
🤖
LMSYS Chatbot Arena
Free
LMSYS Chatbot Arena is an open benchmark platform from UC Berkeley for evaluating large language models through human preference voting. Users chat with two anonymous models simultaneously and vote for the better response. With millions of human votes, Chatbot Arena produces the most reliable ranking of LLM quality through real user preferences rather than automated benchmarks. The Elo-based leaderboard is widely referenced by AI researchers and practitioners for understanding relative model capabilities in real-world conversational settings.
🤖
Open LLM Leaderboard
Free
The Open LLM Leaderboard by HuggingFace tracks and compares open-source large language models across standardised benchmarks including MMLU, ARC, HellaSwag, TruthfulQA, GSM8K, and MATH. It provides an objective, reproducible ranking of hundreds of open-source models updated continuously as new models are released. Researchers and practitioners use the leaderboard to identify the best open-source models for their use case, track progress in the field, and discover newly released models that outperform established ones.
🤖
Together Inference
Freemium
Together AI's inference platform provides the fastest publicly available inference for leading open-source models including Llama 3, Mixtral, Qwen, and DBRX at speeds exceeding 400 tokens per second. The OpenAI-compatible API enables easy migration from proprietary models. Together's custom CUDA kernels and hardware optimisation deliver throughput far exceeding standard GPU deployments. Used by AI startups and enterprises building latency-sensitive applications who need fast, reliable, cost-effective inference for open-source models without managing GPU infrastructure.
🤖
Perplexity Labs Models
Free
Perplexity Labs releases experimental AI models through its labs platform for community testing and feedback. These include Perplexity's own trained models optimised for search-augmented generation and question answering. Perplexity's models are trained with online search capabilities built in, enabling real-time knowledge retrieval as a core model capability rather than an external tool. Used by researchers and AI enthusiasts wanting early access to novel model architectures that tightly integrate search and generation for more accurate, grounded responses.
🤖
Amazon Titan
Paid
Amazon Titan is AWS's family of foundation models available exclusively through Amazon Bedrock. Titan Text models handle text generation, summarisation, and classification. Titan Embeddings produces high-quality vector representations for semantic search and RAG applications. Titan Image Generator creates images from text prompts with watermarking for responsible AI use. Titan models are designed for enterprise reliability, privacy, and AWS integration. Used by enterprises already on AWS who want native foundation models with guaranteed data privacy and seamless integration with AWS services.
🤖
Claude 3 Haiku
Paid
Claude 3 Haiku is Anthropic's fastest and most compact Claude model, designed for near-instant responses in customer-facing applications requiring low latency. Despite its speed, Haiku matches or exceeds competitor models in its class on reasoning and instruction following benchmarks. Its low cost makes it economical for high-volume applications including customer support, content moderation, and data extraction pipelines. Used by enterprises building responsive AI features where speed and cost efficiency are priorities alongside reliable, safe AI behaviour.
🤖
Gemini 1.5 Flash
Freemium
Gemini 1.5 Flash is Google's speed-optimised multimodal model delivering fast responses at low cost while retaining a 1 million token context window. It handles text, images, audio, video, and code with strong performance on most practical tasks. Flash is designed for high-volume, latency-sensitive applications where Gemini 1.5 Pro would be too slow or expensive. Used by developers building applications requiring rapid AI responses across diverse media types — from document Q&A to video understanding — at scale without sacrificing multimodal capabilities.
🤖
OpenAI GPT-4o
Paid
GPT-4o is OpenAI's flagship multimodal model that natively processes and generates text, images, and audio in a single end-to-end model architecture, providing faster and more natural interactions than previous GPT-4 variants while maintaining state-of-the-art performance across reasoning, coding, and knowledge tasks. Its omni capabilities enable real-time voice conversation with natural prosody and visual understanding that opens new interaction modalities for AI applications. Developers and enterprises building sophisticated AI applications use GPT-4o through the OpenAI API for its best-in-class combination of multimodal capability, reasoning performance, and response speed that makes it suitable for the most demanding production AI use cases.
🤖
Claude Anthropic
Freemium
Claude is Anthropic's family of large language models known for their strong reasoning capabilities, nuanced instruction following, long context understanding, and safety-focused design that reduces harmful outputs and hallucinations compared to competing models. The Claude model family provides a range of capability and cost tradeoffs suitable for everything from high-volume automated tasks to complex reasoning and analysis. Enterprises, developers, and researchers building AI applications that require reliable, safe, and capable language model performance use Claude through the Anthropic API for tasks ranging from document analysis to complex multi-step reasoning.
🤖
Google Gemini
Freemium
Google Gemini is Google DeepMind's family of multimodal AI models available through Google AI Studio and Vertex AI that provides state-of-the-art performance across text, code, image, audio, and video understanding tasks with a context window of up to one million tokens in Gemini 1.5 Pro. Its deep integration with Google's search, productivity suite, and cloud infrastructure makes it the natural choice for enterprises building AI applications within the Google ecosystem. Developers and enterprises building AI applications requiring long-context document processing, multimodal understanding, or native integration with Google Workspace and Google Cloud use Gemini for its unique combination of context length and multimodal capability.
🤖
Meta Llama
Free
Meta's Llama family of open-weight large language models provides state-of-the-art performance across reasoning, coding, and knowledge tasks in a freely available model family that can be downloaded, fine-tuned, and deployed without API costs or data privacy concerns. Llama models ranging from 8B to 405B parameters deliver performance competitive with proprietary models while enabling complete customization and on-premises deployment. Researchers, enterprises with data privacy requirements, and developers who want to fine-tune or self-host LLMs use Meta Llama as the foundation model for custom AI applications where control over model behavior, deployment infrastructure, and data handling is essential.
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