LLM models

Mercury by Inception Labs

4.6(5)

A diffusion-based language model that generates text up to 10 times faster than traditional LLMs.

AutoGen

5.0(2)

An open-source framework from Microsoft for creating multi-agent systems based on LLMs.

DolphinGemma

4.5(2)

An artificial intelligence model designed by Google that will attempt to decipher the language of dolphins and learn more about their mode of…

Qwen2.5-Coder

5.0(2)

A family of open-source AI models dedicated to code, available in 6 sizes (0.

ChatALL

5.0(1)

Interact with multiple AI chatbots (+25) in a single query and get better answers.

WeDLM

4.4(7)

A Tencent diffusion language model capable of producing multiple tokens in parallel, with inference speeds up to 10 times faster than traditional…

Devstral2 Mistral

4.5(10)

Code faster with two open-source models (1.

Phi-4

Verified

4.7(3)

A 14-billion-parameter language model that excels at complex reasoning, especially mathematics.

ChatHub

5.0(1)

A Chrome extension that lets you use several ChatBots at once: ChatGPT, Claude, Bard, Bing, etc.

GPT-4o mini

Verified

4.7(3)

OpenAI's low-cost but powerful LLM model: multimodal reasoning, computer coding, long-form writing, etc.

DeepSeek-V2.5

4.5(2)

A powerful LLM model with 236 billion parameters. It performs brilliantly in mathematics, coding and reasoning, with very competitive API pricing ($0.

MiniMax M2

4.4(5)

An open-source LLM with 200 billion parameters and exceptional performance (at low cost).

UniAnimate

5.0(1)

Generate long, consistent videos from a single reference image. This is a unified diffusion model for animating images with human beings

Minigpt-4

4.3(4)

An AI under GPT-4 that can accurately describe an image, create a website, generate a story, etc.

Tongyi DeepResearch

4.5(4)

Automate your web searches with a 30B parameter AI agent: legal analysis, travel, in-depth reports via multi-step reasoning and dynamic tools

Mixtral 8x22B

4.8(6)

Chat for free with many recent LLM models: Mistral Medium, Code Llama 70B, Mixtral 8x7B, Llava 7b, etc.

Nemotron-70B

4.7(3)

A high-performance AI language model customized by NVIDIA.

Llama 3.1 405B

5.0(1)

The free, open-source LLM model from Meta, with performance claimed to be equivalent to OpenAI's GPT-4

QVQ by Qwen

4.5(4)

A 72-billion-parameter multimodal model that excels in visual reasoning.

Leaderboard LLM

4.7(3)

A ranking of the best LLM models based on recognised technical criteria (Elo, MT-Bench or MMLU).

Llama 2

4.7(3)

The new open source language model available for commercial use. Beta version available on request.

Stable Diffusion 3 Medium

4.8(5)

Stability AI's most advanced open-source AI model (2 billion parameters).

Mistral-medium

4.8(4)

One of the most powerful open-source LLM models available today. Currently, the model is only available via an API

Llama 3

4.5(8)

Llama 3 is a high-performance open-source LLM model designed by Meta AI, with 400 billion parameters on the counter