UltraGPT
Every model, one list.
The full catalog UltraGPT draws from — context window, live pricing, capabilities and benchmarks for each model, kept in sync automatically. No guessing which model does what before you switch.
531
Models tracked
65
Providers
392
Reasoning models
346
See images (vision)
2,000K
Largest context window
31
Free to run
Explore and compare AI models
Search hundreds of AI models from leading providers and compare the specifications that matter for your work. Filter by provider and capabilities, then compare context window, reasoning support, input and output modalities, pricing and independent benchmark results where available.
Not sure where to start? Choose a use case above and narrow the catalog to models built for the job.
531 of 531 models
Crouton Alpha is a lightweight model built for scalable coding, agentic, and multimodal workflows. It's ideal for executing agentic sub-tasks, tool use, and document analysis in high-volume, repeated work. Developers...
GLM-5.3-FlashX-Legal (GLM-Legal) is Z.ai's legal-domain variant of GLM-5.3-FlashX, a native multimodal model that accepts text, image, and video input and produces text. It adds legal-domain mid-training and post-training with reinforcement...
Particle-1 is a token-efficient agentic coding model from Particle AI, post-trained on DeepSeek V4.1 Flash. It is suited for software engineering work such as terminal workflows, codebase changes, and data...
GPT-6.1 Terra is the mid-tier model in OpenAI's GPT-6.1 series, priced below GPT-6.1 Sol. It is suited for everyday agentic, coding, and writing work, is particularly strong at understanding screenshots...
GPT-6.1 Terra is the mid-tier model in OpenAI's GPT-6.1 series, priced below GPT-6.1 Sol. It is suited for everyday agentic, coding, and writing work, is particularly strong at understanding screenshots...
GPT-6.1 Terra Pro is the same underlying model as GPT-6.1 Terra, served with reasoning.mode set to pro for higher-quality responses on complex tasks. Cost note: pro mode spends far more...
GPT-6.1 Terra Pro is the same underlying model as GPT-6.1 Terra, served with reasoning.mode set to pro for higher-quality responses on complex tasks. Cost note: pro mode spends far more...
Kimi K3.1 Pro is Moonshot AI's flagship open-weight multimodal reasoning model and a post-trained successor to Kimi K3, with a 1M-token context window. It is built for long-horizon coding, 3D...
Design Arena — ranked #1 across 11 categories
Flint is an open-weight mixture-of-experts model from Recursive, with about 2.5B active parameters out of 40B total, built by Recursive's autoresearch system Eureka Machine for coding agents. It is suited...
MiniMax-M3.1-Flash is a multimodal model from MiniMax. Compared with M3, it delivers significant improvements across multiple benchmarks, particularly in complex agentic tasks, coding, and cybersecurity. It is built for complex,...
MiniMax-M3.1-Flash is a multimodal model from MiniMax. Compared with M3, it delivers significant improvements across multiple benchmarks, particularly in complex agentic tasks, coding, and cybersecurity. It is built for complex,...
DeepSeek V4 Flash is an efficiency-optimized Mixture-of-Experts model from DeepSeek with 284B total parameters and 13B activated parameters, supporting a 1M-token context window. It is designed for fast inference and...
Design Arena — ranked #47 across 8 categories
DeepSeek V4 Flash Vision Exp is an experimental vision-enabled version of DeepSeek V4 Flash 0731 from DeepSeek, adding image understanding while matching the base model on text capabilities including agents,...
DeepSeek V4.1 Flash 1017 is the October 17, 2026 snapshot of DeepSeek V4.1 Flash, served by DeepSeek under the deepseek-flash API model name. V4.1 Flash is a sparse mixture-of-experts model...
Design Arena — ranked #6 across 9 categories
Glimpse is a multimodal visual reasoning model from Elorian, built for visual understanding and complex visual reasoning in the physical world. It reasons directly over images and video rather than...
Mercury 3 is a diffusion-based language model from Inception, and the lab's most capable model to date. It is suited for agentic workflows and real-time applications where latency compounds, such...
Design Arena — ranked #41 across 6 categories
Ternary Bonsai 2 9B is PrismML’s compact, open-weight language model for conversational AI, reasoning, and tool use. Built on Qwen3.5, it uses ternary weights packed into 2 bits for efficient...
Beam 1.0 is Reflection's open-weight reasoning model for coding and long-horizon agentic workflows. It uses a sparse mixture-of-experts architecture with 501B total and 23B active parameters per token, and a...
Design Arena — ranked #52 across 3 categories
Xing4.0-29B-A4B is an open-weight mixture-of-experts language model from China Telecom's AI team (TeleAI) and the successor to the TeleChat series. It has 29B total parameters with 4B active per token...
Muse Spark Open is a coding-optimized open-weight model. Trained specifically for agentic coding workflows, it delivers higher first-attempt accuracy, reliable tool calling, and deep codebase understanding across the repos developers...
Design Arena — ranked #13 across 4 categories
Moondream 3 Preview is a vision-language model from Moondream with 9B total parameters and 2B active (mixture-of-experts). It handles visual question answering, captioning, OCR, and visual reasoning over images, with...
Moondream 3.1 is a vision-language model from Moondream, the successor to Moondream 3 Preview, with 9B total parameters and 2B active (mixture-of-experts). It handles visual question answering, captioning, OCR, and...
NVIDIA Nemotron 3.5 Super VL is an open multimodal mixture-of-experts model from NVIDIA, with 12B active parameters out of 120B total. It takes text, images, documents, GUI screenshots, and video...
Design Arena — ranked #82 across 5 categories
NVIDIA Nemotron 3.5 Super VL is an open multimodal mixture-of-experts model from NVIDIA, with 12B active parameters out of 120B total. It takes text, images, documents, GUI screenshots, and video...
Design Arena — ranked #82 across 5 categories
How to choose the right AI model
The best AI model depends on what you are trying to accomplish. A model that performs well on complex reasoning tasks may not be the fastest option for everyday questions, while a model with a very large context window may be more useful when working with long documents or codebases. Here are the main factors worth considering.
Start with the task
- For coding, look at coding performance, reasoning ability and tool support.
- For research and complex analysis, reasoning capability and a larger context window can be more important.
- For writing, brainstorming and everyday conversations, response quality and speed may matter more than maximum reasoning performance.
- For images, screenshots and visual documents, choose a model that supports vision input.
Consider context window
The context window determines how much information a model can work with at once. Larger context windows can be useful for long documents, large codebases, lengthy conversations and research involving multiple sources. A larger context window, however, does not automatically mean a model is better at every task.
Look at reasoning capabilities
Reasoning models spend additional computation working through difficult problems before producing an answer. They can be useful for programming, mathematics, planning, research and other multi-step tasks, while simpler tasks may benefit more from faster models.
Check supported modalities
AI models can support different types of input and output. Depending on the model, you may be able to work with text, images, files, audio or video. Check each model's capabilities before choosing it for a multimodal workflow.
Balance capability and speed
The most capable model is not always the most practical model. For quick questions, rewriting, classification or simple workflows, a faster model may provide a better experience. For difficult problems, a more capable reasoning model may be worth the additional processing time.
Compare models instead of relying on one
Different models have different strengths. Comparing several models on the same task is often the easiest way to discover which one works best for your workflow.