| live |
GPT-4o |
https://chat.openai.com/ |
gpt-4o |
 OpenAI |
OpenAI API |
GPT-4o (“o” for “omni”) is a step towards much more natural human-computer interaction—it accepts as input any combination of text, audio, and image and generates any combination of text, audio, and image outputs. |
May 2024 |
May 2024 |
128k context window, October 2023 training data |
Proprietary |
| live |
GPT-4o |
https://chat.openai.com/ |
gpt-4o, same as GPT4 Omni |
.webp) OpenAI |
OpenAI API |
GPT-4o (“o” for “omni”) is a step towards much more natural human-computer interaction—it accepts as input any combination of text, audio, and image and generates any combination of text, audio, and image outputs. |
May 2024 |
May 2024 |
128k context window, October 2023 training data |
Proprietary |
| live |
GPT-4o (Vision) |
https://chat.openai.com/ |
gpt-4o, with vision/image read |
.webp) OpenAI |
OpenAI API |
GPT-4o (“o” for “omni”) is a step towards much more natural human-computer interaction—it accepts as input any combination of text, audio, and image and generates any combination of text, audio, and image outputs. |
May 2024 |
May 2024 |
128k context window, October 2023 training data |
Proprietary |
| live |
OpenAI o1 |
https://chat.openai.com/ |
o1 |
 OpenAI |
ChatGPT Plus/Pro |
The o1 series of large language models are trained with reinforcement learning to perform complex reasoning. |
December 2024 |
December 2024 |
200k context window, October 2023 training data |
Proprietary |
| live |
OpenAI o1 (Vision) |
https://chat.openai.com/ |
o1-vision |
 OpenAI |
ChatGPT Plus/Pro |
The o1 series of large language models are trained with reinforcement learning to perform complex reasoning. |
December 2024 |
December 2024 |
200k context window, October 2023 training data |
Proprietary |
| live |
OpenAI o1 Pro |
https://chat.openai.com/ |
o1-pro |
 OpenAI |
ChatGPT Plus/Pro |
The o1 series of large language models are trained with reinforcement learning to perform complex reasoning. |
December 2024 |
December 2024 |
200k context window, October 2023 training data |
Proprietary |
| live |
OpenAI o1 Pro (Vision) |
https://chat.openai.com/ |
o1-pro-vision |
 OpenAI |
ChatGPT Plus/Pro |
The o1 series of large language models are trained with reinforcement learning to perform complex reasoning. |
December 2024 |
December 2024 |
200k context window, October 2023 training data |
Proprietary |
| live |
OpenAI o3 mini |
https://chat.openai.com/ |
o3-mini |
 OpenAI |
OpenAI API |
o3-mini is a powerful and fast model that advances the boundaries of what small models can achieve, delivering exceptional STEM capabilities—with particular strength in science, math, and coding—all while maintaining the low cost and reduced latency of OpenAI o1-mini. |
January 2025 |
January 2025 |
200k context window, October 2023 training data |
Proprietary |
| live |
OpenAI o3 mini high |
https://chat.openai.com/ |
o3-mini-high |
 OpenAI |
ChatGPT Plus/Pro |
o3-mini is a powerful and fast model that advances the boundaries of what small models can achieve, delivering exceptional STEM capabilities—with particular strength in science, math, and coding—all while maintaining the low cost and reduced latency of OpenAI o1-mini. |
January 2025 |
January 2025 |
200k context window, October 2023 training data |
Proprietary |
| live |
GPT4.5 Preview |
https://chat.openai.com/ |
gpt-4.5-preview |
 OpenAI |
OpenAI API |
GPT 4.5 Preview is a research preview of OpenAI's strongest GPT model. GPT‑4.5 is OpenAI's largest and best model for chat yet |
February 2025 |
February 2025 |
128k context window, October 2023 training data |
Proprietary |
| live |
OpenAI o3 |
https://chat.openai.com/ |
o3 |
 OpenAI |
OpenAI API |
OpenAI o3 is our OpenAI's powerful reasoning model that pushes the frontier across coding, math, science, visual perception, and more. |
April 2025 |
April 2025 |
200k context window, June 2024 Knowledge Cut-off |
Proprietary |
| live |
OpenAI o4 mini |
https://chat.openai.com/ |
o4-mini |
 OpenAI |
OpenAI API |
OpenAI o4-mini is a smaller model optimized for fast, cost-efficient reasoning—it achieves remarkable performance for its size and cost, particularly in math, coding, and visual tasks. |
April 2025 |
April 2025 |
200k context window, June 2024 Knowledge Cut-off |
Proprietary |
| live |
OpenAI o4 mini high |
https://chat.openai.com/ |
o4-mini-high |
 OpenAI |
ChatGPT Plus/Pro |
OpenAI o4-mini is a smaller model optimized for fast, cost-efficient reasoning—it achieves remarkable performance for its size and cost, particularly in math, coding, and visual tasks. |
April 2025 |
April 2025 |
200k context window, June 2024 Knowledge Cut-off |
Proprietary |
| live |
DeepSeek R1 |
|
deepseek-r1 |
 DeepSeek |
self-hosting |
DeepSeek's first generation reasoning models with comparable performance to OpenAI-o1. |
January 2025 |
|
|
MIT |
| live |
DeepSeek V3 |
|
deepseek-v3-0324 |
 DeepSeek |
self-hosting |
A strong Mixture-of-Experts (MoE) language model with 671B total parameters with 37B activated for each token. |
December 2024 |
March 2025 |
|
MIT |
| live |
Gemini 2.5 Pro Exp. |
https://gemini.google.com/ |
gemini-2.5-pro-exp-03-25 |
 Google |
https://gemini.google.com/ |
Gemini Pro is the strongest model quality, especially for code & world knowledge; 2M long context. |
December 2023 |
February 2025 |
2M context window |
Proprietary |
| live |
Gemini 2.0 Flash Thinking Exp. |
https://gemini.google.com/ |
gemini-2.0-flash-thinking-exp |
 Google |
https://gemini.google.com/ |
Gemini Thinking provides stronger reasoning capabilities and includes the thinking process in responses. |
December 2023 |
February 2025 |
1M context window |
Proprietary |
| live |
Gemini Advanced (Vision) |
https://gemini.google.com/ |
gemini-2.0-pro, with vision/image read |
.webp) Google |
https://gemini.google.com/ |
Gemini is a family of multimodal large language models developed by Google DeepMind, serving as the successor to LaMDA and PaLM 2. |
December 2023 |
February 2025 |
1M context window |
Proprietary |
| live |
Bing Copilot |
https://copilot.microsoft.com/ |
sydney, gpt-4-turbo (precise mode) |
 Microsoft |
https://copilot.microsoft.com/ |
Microsoft Copilot is a large language model-based chatbot developed by Microsoft. It was launched as Bing Chat on February 7, 2023, as a built-in extension for Microsoft Edge, and is the indirect successor to Cortana. Over the course of 2023, Microsoft began to unify the Copilot branding across its various chatbot products. Copilot's conversational interface style resembles that of ChatGPT. It is able to cite sources, create poems and songs, and understand and communicate in numerous languages and even certain dialects. |
February 2023 |
March 2024 |
unknown |
Proprietary |
| live |
Claude-3 Opus |
https://claude.ai/ |
claude-3-opus |
 Anthropic |
https://claude.ai/ |
The Claude 3 family includes three state-of-the-art models in ascending order of capability: Claude 3 Haiku, Claude 3 Sonnet, and Claude 3 Opus. Each successive model offers increasingly powerful performance, allowing users to select the optimal balance of intelligence, speed, and cost for their specific application. |
March 2024 |
April 2024 |
200k context window (Max at 1M context window) |
Proprietary |
| live |
Claude 3.7 Sonnet Extended |
https://claude.ai/ |
claude-3.7-sonnet-extended-thinking |
 Anthropic |
https://claude.ai/ |
Claude 3.7 Sonnet, Anthropic's most intelligent model to date and the first hybrid reasoning model on the market. |
Feb 2025 |
Feb 2025 |
200k context window |
Proprietary |
| live |
Claude-3.7 (Vision) |
https://claude.ai/ |
claude-3.7-sonnet-extended-thinking, with vision/image read |
.webp) Anthropic |
https://claude.ai/ |
Claude 3.7 Sonnet, Anthropic's most intelligent model to date and the first hybrid reasoning model on the market. |
Feb 2025 |
Feb 2025 |
200k context window |
Proprietary |
| live |
Llama-3.3 |
https://meta.ai/ |
llama-3.3-70b |
 Meta |
https://meta.ai/ |
The Meta Llama 3.3 multilingual large language model (LLM) is a pretrained and instruction tuned generative model in 70B (text in/text out). The Llama 3.3 instruction tuned text only model is optimized for multilingual dialogue use cases and outperform many of the available open source and closed chat models on common industry benchmarks. |
April 2024 |
December 2024 |
70 billion |
Meta License |
| live |
Llama 4 Maverick |
https://meta.ai/ |
llama-4-Maverick-17B-128E-Instruct-FP8 |
 Meta |
https://meta.ai/ |
The Llama 4 herd: The beginning of a new era of natively multimodal AI innovation/ |
April 2025 |
April 2025 |
17B active param, 128 Experts |
Meta License |
| live |
Llama-3.2 (Vision) |
https://meta.ai/ |
llama-3.2-90b-vision |
.webp) Meta |
https://meta.ai/ |
Llama 3.2: Revolutionizing edge AI and vision with open, customizable models |
April 2024 |
September 2024 |
90 billion, 128k context window |
Meta License |
| live |
Grok-3 |
https://x.com/grok |
grok-3 |
 xAI |
https://x.com/grok |
Grok-3 is xAI's most advanced model yet; blending strong reasoning with extensive pretraining knowledge. |
February 2025 |
February 2025 |
unknown |
Apache 2.0 |
| live |
Grok-3 Think |
https://x.com/grok |
grok-3-think |
 xAI |
https://x.com/grok |
Grok-3 is xAI's most advanced model yet; blending strong reasoning with extensive pretraining knowledge |
February 2025 |
February 2025 |
unknown |
Apache 2.0 |
| live |
Cohere |
https://coral.cohere.com/ |
c4ai-aya-23-35b |
 Cohere |
Cohere API |
Coral is an AI chat interface/toolkit developed by Cohere, that can search the web to cite sources. Coral can help research a new industry, start a draft, or debug code. |
July 2023 |
May 2024 |
128k context window |
CC-by-NC-4.0 |
| live |
Solar |
https://console.upstage.ai/ |
solar-pro-preview |
 UpStage |
Upstage API |
Upstage said Solar stands for Specialized and Optimized LLM and Applications with Reliability. The language model is based on the company's generative AI capability, which took first place on Hugging Face's Open LLM Leaderboard competition with a score of 72.3 in August. |
July 2023 |
September 2024 |
unknown |
Apache 2.0 |
| live |
Phi |
https://huggingface.co/chat/settings/microsoft/Phi-3-mini-4k-instruct |
phi-4 |
 Microsoft |
self-hosting |
Phi is a small language model capable of common-sense reasoning and language understanding. It showcases “state-of-the-art performance” among language models with less than 13 billion parameters. |
September 2023 |
January 2025 |
3 billion |
MIT |