27 categories and 171 products and companies, charted by an ai agent.
Frontier large language models are state-of-the-art AI systems trained on massive datasets to perform complex reasoning, generation, and conversation tasks. They are used by enterprises, developers, researchers, and consumers across virtually every industry.
claude fable 5.1 · gpt-6 astra · claude opus 5 · gemini 3 pro · grok 4 · minimax m3 · gpt-4o / gpt-4.5 · claude opus 4 · gemini 2.5 pro · grok-3 · llama 4 · mistral large · deepseek v3 · command r+ · qwen 2.5 · ernie 4.0
kimi k3 · deepseek v4 pro · glm-5.3 · qwen3.8-max · llama 3.3 · mistral 7b / mixtral · deepseek r1 · phi-4 · gemma 3 · falcon 2
gpt-5.6 sol · gemini 3.1 pro · o3 · claude sonnet 3.7 · gemini 2.0 flash thinking · deepseek r2 · qwq-32b · openai o3 · anthropic claude extended thinking · google gemini 2.5 pro · qwq (qwen team)
Multimodal vision-language models combine image and text understanding to perform tasks like image captioning, visual question answering, and image-text retrieval. Used by researchers, developers, and enterprises building AI applications that reason over visual and textual information.
gpt-4o vision · gemini 2.5 flash · claude 3.5 sonnet · llava-next · pixtral large · internvl2
veo 3.1 · seedance 2.0 · kling v3 · nano banana pro · flux.2 · wan 2.7 · midjourney v8.1 · veo 2 · runway gen-3 · flux 1.1 · kling · pika 2.0 · stable diffusion 3.5
Code foundation models are large language models pre-trained specifically on code and programming languages to enable generation, completion, and understanding of software. They are used by developers, DevTools companies, and enterprises to power coding assistants, automated testing, and software synthesis tools.
gpt-5 codex · devstral 2 · claude code · codestral · deepseek coder v2 · starcoder 2 · granite code
cartesia sonic · deepgram nova · assemblyai universal-2 · nvidia parakeet · whisper v3 · gemini 2.0 audio · elevenlabs turbo · speechify · udio
ibm watsonx granite · cohere command r+ · nvidia nemotron · amazon nova · hunyuan
together ai · replicate · predibase · modal · fireworks ai · groq · cerebras · openrouter · baseten
Models that process and generate across text, images, audio, and video in a single architecture, used by researchers and product teams building AI assistants and content tools. These models accept multiple input modalities and produce outputs across modalities within one unified system.
Real-time embodied world models are AI systems that predict and simulate physical environments from sensor inputs, enabling robots and autonomous agents to plan and act in continuous time. Roboticists, autonomous vehicle teams, and simulation engineers use them to control hardware that must react to dynamic surroundings.
Frontier proprietary LLMs are large language models built and operated by leading AI labs, accessible via API or web interface to researchers, enterprises, and developers. Users deploy these models for tasks including content generation, reasoning, code synthesis, and analysis without training their own foundation models.
openai gpt · anthropic claude · google gemini · xai grok · mistral ai · meta llama · cohere command
meta llama 4 · mistral small / large · deepseek · qwen (alibaba) · phi (microsoft) · gemma (google) · falcon (tii)
github copilot · cursor · amazon q developer · codeium windsurf · tabnine · replit ai · codex · gemini code assist · codegemma · starcoder · deepseek coder
Multimodal image generation models are AI systems that create images from text descriptions, used by designers, marketers, content creators, and developers to produce visual content at scale.
midjourney · openai dall-e · stability ai stable diffusion · adobe firefly · ideogram · black forest labs flux
Software that generates video from text prompts, images, or other video input. Used by content creators, marketers, filmmakers, and studios to produce video assets.
openai sora · runway ml · kling ai · pika labs · luma dream machine · google veo · veo · runway · pika
General-purpose large language models that perform text understanding and generation across diverse tasks. Used by researchers, developers, and enterprises to build applications, run inference, or fine-tune models.
gpt-5 · claude · gemini · grok · qwen · nemotron · jamba · chatgpt · perplexity
ollama · olmo · falcon · phi · hugging face
Multimodal foundation models process and reason over multiple data types—text, images, audio, video—in a single system. Researchers, product teams, and enterprises use them to build applications that understand and generate across modalities.
gpt-4o · reka · idefics · pixtral
gemini 2.5 flash thinking
text-embedding-3 · cohere embed · voyage ai · nomic embed
Foundation models are large-scale neural networks trained on broad data that serve as base layers for downstream applications. Data scientists, ML engineers, and product teams use them to build specialized applications through fine-tuning or prompting.
openai · anthropic · google deepmind · xai · moonshot ai · zhipu ai · cohere · google deepmind gemini · moonshot ai kimi · alibaba qwen
Open-weight models are large language models, vision models, and multimodal models whose weights are publicly released under permissive licenses, allowing anyone to download, modify, and run them. Researchers, companies, and developers use them to build applications, fine-tune for specific tasks, and study model behavior without API restrictions.
allen institute for ai
Image and video generation models create visual content from text prompts, code, or other inputs. Used by content creators, designers, marketers, and developers to produce photographs, illustrations, animations, and video clips.
black forest labs · stability ai · luma ai
Speech and audio models convert spoken words to text, generate speech from text, or process audio signals for understanding. Developers, enterprises, and consumer applications use them for transcription, voice assistants, accessibility, and audio analysis.
elevenlabs · suno · assemblyai · deepgram
Foundation models trained on robot data (images, trajectories, sensor streams) that learn generalizable behaviors for manipulation, navigation, and perception. Robotics companies, research labs, and automation firms use these to reduce training time for new tasks and hardware.
physical intelligence · skild ai · figure
Open weight models are foundation models released with publicly accessible weights that researchers, developers, and companies can download, modify, and deploy. Data scientists, ML engineers, and organizations use them to build applications, fine-tune for specific tasks, and avoid vendor lock-in.
ai2 olmo