Models

Imagen 4 is a Google text-to-image model announced on 20 May 2025 that outputs images at up to 2K resolution with improved fine detail and typography, shipping in the Gemini app, Whisk, Vertex AI and Workspace before its API models were retired in 2026.

Generative modelsCreative toolsImage generation

Kling 2.1 is a Kuaishou video generation model released at the end of May 2025 in three tiers — 720p Standard, 1080p High Quality and 1080p Master — with claimed gains in motion, physical simulation and prompt responsiveness. It was superseded by Kling 2.5 Turbo.[^1][^2][^3]

ChinaMultimodalGenerative models

π0.5 is a vision-language-action model from Physical Intelligence, released in April 2025, built to generalize to unfamiliar homes and trained by co-training on many kinds of robot and web data.

Deep learningFoundation modelsMultimodal

ByteDance Seed's GUI agent model from April 2025 and the successor to UI-TARS. Its 7B version is open-source under Apache-2.0, and the family reports state-of-the-art results on several screen-control benchmarks.

Open sourceChinaAgents & tool use

Kling 2.0 is Kuaishou's AI video generator, launched globally in April 2025 and billed by the company as the world's most powerful, with a multimodal editing feature and the Multi-modal Visual Language concept.

ChinaMultimodalGenerative models

Seedream 3.0 is ByteDance Seed's Chinese–English image-generation foundation model of 2025, improving on Seedream 2.0 in prompt alignment, Chinese text rendering and native 2K output. Seedream 4.0 succeeded it in September 2025.

ChinaGenerative modelsCreative tools

A speech-to-speech foundation model in Amazon's Nova family, announced on 8 April 2025, which handles recognition, understanding and speech generation in a single model rather than a chain of separate ones.

Agents & tool useMultimodalGenerative models

Midjourney V7 is an image generation model released as an alpha to Midjourney's community on 4 April 2025, the first of the company's models to have personalization on by default and the first to add a cheap, fast Draft Mode.

Consumer appsGenerative modelsCreative tools

Kimi-VL is an open-source vision-language model from Moonshot AI, released in April 2025. It pairs a 16B-parameter mixture-of-experts language model with a native-resolution vision encoder, so only about 3B parameters are active at a time.

Open sourceChinaMultimodal

Runway Gen-4 is a video generation model released on 31 March 2025 that keeps characters, locations and objects consistent across shots, working from reference images and text instructions rather than per-project training.

MultimodalGenerative modelsCreative tools

Isaac GR00T N1 is an open foundation model for humanoid robots that NVIDIA announced on 18 March 2025. It is designed to be post-trained for a specific robot or task, and it pairs a planning model with quick motion control.

Open sourceFoundation modelsRobotics

The largest model in Ai2's OLMo 2 family: a 32-billion-parameter language model released with its data, code, weights and training details, and reported to outperform GPT-3.5-Turbo and GPT-4o mini on academic benchmarks.

Large language modelsOpen sourceData & datasets

Gemini Robotics is a vision-language-action model from Google DeepMind, built on Gemini 2.0, that outputs physical actions to control robots. It was announced on 12 March 2025 alongside Gemini Robotics-ER.

MultimodalRoboticsComputer vision

Mistral OCR is a document-understanding API released by Mistral AI on 6 March 2025 that reads text, tables, equations and embedded media across thousands of scripts and languages, and was later superseded by newer OCR versions.

EnterpriseData & datasetsComputer vision

Cohere's first vision-language model, released in March 2025 as open-weight 8B and 32B versions. It handles images and text in 23 languages and, in Cohere's comparisons, beats several much larger models, under a non-commercial licence.

Large language modelsOpen sourceMultimodal

Wan2.1 is Alibaba's open-source video generation model family, released on 25 February 2025 under Apache 2.0. It creates videos from text or images and can render Chinese and English text.

Open sourceChinaGenerative models

Step-Audio is an open family of speech-interaction models released by StepFun on 17 February 2025. It includes Step-Audio-Chat, with 130 billion parameters, and Step-Audio-TTS-3B.

Open sourceChinaMultimodal

Step-Video-T2V is an open text-to-video model with 30 billion parameters. StepFun released its weights, code, a Turbo variant and a technical report under the MIT license on 17 February 2025.

Open sourceChinaGenerative models

Hunyuan3D 2.0 is Tencent's large-scale system for generating high-resolution textured 3D assets: a shape diffusion transformer plus a texture model, released with code and pretrained models on 21 January 2025 alongside its technical report.

Open sourceChinaGenerative models

UI-TARS is a GUI agent model from ByteDance Seed: it takes only screenshots as input and outputs keyboard and mouse actions, reported state-of-the-art results on more than 10 benchmarks, and was released with open weights under Apache-2.0.

Open sourceChinaAgents & tool use

NVIDIA Cosmos is a platform of world foundation models that generate physics-based video to train robots and autonomous vehicles, released as open-weight models in January 2025.

Open sourceFoundation modelsRobotics

Qwen2.5-VL is Alibaba's vision-language model family released in January 2025 in 3B, 7B, 32B and 72B sizes. It reads charts and documents, localises objects, works as a visual agent and understands videos over an hour long.

Open sourceChinaAgents & tool use

Veo 2 is a video-generation model Google announced on 16 December 2024, emphasising real-world physics and human movement, with outputs up to 4K that can extend to minutes.

MultimodalGenerative modelsCreative tools

A family of open mixture-of-experts vision-language models from DeepSeek, released in three sizes from 1.0B to 4.5B activated parameters for document, chart and OCR understanding, succeeding DeepSeek-VL.

Open sourceChinaMultimodal

Phi-4 is Microsoft's 14-billion-parameter dense decoder-only language model, built with a data-quality-first recipe heavy on synthetic data and released with full weights under the MIT licence in January 2025.

Large language modelsOpen sourceReasoning

Genie 2 is a foundation world model from Google DeepMind, announced on 4 December 2024, that turns a single prompt image into a playable 3D environment that a person or an AI agent can explore with keyboard and mouse controls.

Agents & tool useGenerative modelsSimulation

HunyuanVideo is Tencent's open-source text-to-video generation model, with over 13 billion parameters, released in December 2024. Running it locally needs at least 60GB of GPU memory.

Open sourceChinaMultimodal

Amazon Nova is Amazon's family of foundation models, introduced on 3 December 2024 and offered exclusively in Amazon Bedrock. It spans text understanding models from Micro to Pro, plus Canvas for images and Reel for video, and was succeeded by Nova 2 in December 2025.

Large language modelsEnterpriseMultimodal

Kling 1.6 is a December 2024 upgrade to Kuaishou's Kling video generation model, with better handling of motion and camera instructions in prompts and improved visual quality; it was succeeded by Kling 2.0 in April 2025.

ChinaMultimodalGenerative models

OLMo 2 is the second generation of the Allen Institute for AI's fully open language model family, released in 7B and 13B sizes with weights, data, code and training recipes all public.

Large language modelsOpen sourceData & datasets

Pixtral Large is Mistral AI's open-weight multimodal model announced on 18 November 2024, built on Mistral Large 2, combining a 123B-parameter decoder with a 1B vision encoder and a 128K-token context window. It was retired in May 2026.

EnterpriseMultimodalComputer vision

SmolLM2 is Hugging Face's second family of small language models, released in November 2024 at 135M, 360M and 1.7B parameters, with open weights and a data-centric training recipe.

Large language modelsOpen sourceConsumer apps

π0 is a generalist robot policy from Physical Intelligence, built by adapting a 3-billion-parameter vision-language model with flow matching so one model can output fast, continuous motor commands for dexterous manipulation.

Deep learningFoundation modelsMultimodal

A family of text-to-image models from Stability AI, announced in October 2024: the 8.1-billion-parameter Large, a 4-step Turbo variant and the 2.5-billion-parameter Medium, released under a community licence.

Open sourceGenerative modelsCreative tools

Movie Gen is a set of media foundation models Meta premiered on 4 October 2024. It covers video generation, personalised video, precise video editing and audio generation, and was not released publicly.

Foundation modelsGenerative modelsCreative tools

A closed text-to-image model from Black Forest Labs, released on 2 October 2024 and available only through an API. Black Forest Labs said it generates six times faster than its predecessor, and it reached the top of the Artificial Analysis image arena leaderboard.

Generative modelsCreative toolsImage generation

Molmo is a family of open vision-language models released by Ai2 on 25 September 2024 that can point at objects in images, with weights, code, data and evaluations published openly.

Open sourceMultimodalData & datasets

Llama 3.2, announced on 25 September 2024, added vision models at 11 billion and 90 billion parameters and small text-only models at 1 billion and 3 billion for edge and mobile devices.

Large language modelsOpen sourceMultimodal

Pixtral 12B is Mistral AI's first multimodal model, pairing a 12B language decoder with a 400M vision encoder. Released openly under Apache 2.0 in September 2024, it reads images at native resolution and was retired at the end of 2025.

Open sourceMultimodalComputer vision

NVLM is a family of multimodal large language models from NVIDIA, described in a paper posted to arXiv in September 2024; its flagship NVLM-D-72B was later released as open weights for non-commercial use.

Large language modelsOpen sourceMultimodal

Kling 1.5 is a September 2024 version of Kuaishou's video generation model, adding 1080p output, a Motion Brush, and quality improvements before being succeeded by Kling 1.6 in December 2024.

ChinaMultimodalGenerative models

Qwen2-VL is Alibaba's vision-language model released on 29 August 2024. It understands videos longer than 20 minutes, reads text in many languages, and can act as an agent on phones and robots.

Open sourceChinaAgents & tool use

CogVideoX is an open-source family of text-to-video models from Zhipu AI and Tsinghua University, released starting in August 2024 and derived from the earlier open-source CogVideo.

Open sourceChinaGenerative models

FLUX.1 is a family of 12-billion-parameter text-to-image models released by Black Forest Labs on 1 August 2024 in pro, dev and schnell variants. The dev version is open-weight, and the variants carry different licenses.

Open sourceGenerative modelsCreative tools

SAM 2 is Meta's 2024 successor to the Segment Anything Model. It segments objects in both images and videos in real time, and its release included the SA-V video dataset.

Open sourceFoundation modelsData & datasets

SmolLM is a family of small language models with 135M, 360M and 1.7B parameters, released openly by Hugging Face in July 2024 for on-device use and trained on a purpose-built corpus of generated textbooks, code and filtered web text.

Large language modelsOpen sourceConsumer apps

Kolors is Kuaishou's open-weight text-to-image model, released on 6 July 2024, built on latent diffusion and tuned for both Chinese and English prompts.

Open sourceChinaGenerative models

ESM3 is a protein generative model announced by EvolutionaryScale in June 2024 that reasons over sequence, structure and function together, and generated a fluorescent protein estimated to sit over 500 million years of evolution away from known ones.

Open sourceFoundation modelsGenerative models

Claude 3.5 Sonnet, released on 20 June 2024, was the first model of the Claude 3.5 family. It ran twice as fast as Claude 3 Opus at the mid-tier price and powered the new Artifacts and Projects features.

Large language modelsEnterpriseConsumer apps

Runway's 2024 video generation model, trained jointly on video and images, with finer temporal control such as keyframing; later extended with an API, video-to-video and Act-One performance capture.

MultimodalGenerative modelsCreative tools

Nemotron-4 340B is a family of open models from NVIDIA, released on 14 June 2024 as base, instruct and reward models. It was built to generate synthetic data for training other language models, and it fits on a single eight-GPU DGX H100.

Large language modelsOpen sourceEnterprise

Veo is a generative video model from Google DeepMind, announced at Google I/O on 14 May 2024. Google described it as its most capable video model, able to produce 1080p clips longer than a minute and later succeeded by Veo 2.

MultimodalGenerative modelsCreative tools

Google's text-to-image model announced at I/O in May 2024, which Google described as its highest-quality image generator, with fewer visual artifacts and improved text rendering.

Generative modelsCreative toolsImage generation

Hunyuan-DiT is Tencent's open text-to-image diffusion transformer, built for fine-grained English and Chinese prompt understanding. Its 2024 report and weights were released publicly, and human evaluations placed it ahead of several open models but behind leading closed ones.

Open sourceChinaGenerative models

Octo is an open-source generalist robot policy: a transformer-based diffusion model trained on 800,000 robot trajectories that can drive a range of robots out of the box and be fine-tuned to new setups.

Open sourceDeep learningFoundation models

OpenELM is a family of small open language models from Apple, published in four sizes from 270M to 3B parameters with its training framework, logs, checkpoints and MLX conversion code.

Large language modelsResearch historyOpen source

Grok-1.5 is xAI's upgrade to Grok-1, with better reasoning and a 128,000-token context window. Its companion Grok-1.5V was xAI's first multimodal model. Neither model's weights were released publicly.

Large language modelsMultimodalReasoning

Step-1V is a roughly 100-billion-parameter multimodal model from StepFun, unveiled in March 2024, that reads text, data and charts in images and supports video understanding; it ranked first on OpenCompass at launch and was succeeded by Step-1.5V.

Large language modelsChinaMultimodal