GPT-5.6 Luna
GPT-5.6 Luna is the fast, cost-efficient model in the GPT-5.6 family, intended for high-volume, cost-sensitive workloads. OpenAI announced a limited preview on June 26, 2026, initially for selected…
MODEL DIRECTORY
New releases, in one place. Explore models by their publication date.
GPT-5.6 Luna is the fast, cost-efficient model in the GPT-5.6 family, intended for high-volume, cost-sensitive workloads. OpenAI announced a limited preview on June 26, 2026, initially for selected…
GPT-5.6 Terra is OpenAI’s balanced GPT-5.6 tier for workloads that need a combination of intelligence and cost efficiency. OpenAI describes it as roughly corresponding to the mini tier in earlier GPT-…
GPT-5.6 Sol is OpenAI’s flagship GPT-5.6 model for complex professional work. OpenAI announced a limited preview on June 26, 2026, initially available to selected trusted partners through the API and…
Qwen-Image-2.0-Pro is a full-capability image generation and editing model served through Alibaba Cloud Model Studio. This snapshot was listed in Alibaba Cloud’s release announcement on June 25, 2026;…
GPT-5.5 Instant is an everyday conversational model used in ChatGPT. OpenAI began rolling it out as ChatGPT’s default model on May 7, 2026. The release identified here is an update announced on June 2…
Qwen-AgentWorld-35B-A3B is an open-weight language world model designed to simulate environments for AI agents. Given an agent’s action and interaction history, it predicts the next environment…
Doubao-Seed-2.1-Turbo is a deep-reasoning model positioned as a lower-cost, lower-latency Seed 2.1 option for production at scale. Its documented capabilities include text generation, multimodal…
Doubao-Seed-2.1-Pro is ByteDance’s flagship reasoning model for coding, agent workflows and multimodal understanding. The Seed 2.1 series was announced on June 23, 2026; Volcano Engine’s release notes…
Unlimited-OCR is Baidu’s open-weight vision-language model for OCR and long-document parsing. The original model was released on June 22, 2026. It is distinct from Unlimited-OCR 2, released later.…
Seed-2.1-Pro-Preview is an early-access foundation model released on June 19, 2026 for community testing in Arena AI’s Code Arena. The announcement said availability in Feishu Spark and Coze was…
Fun-ASR-Flash is a non-real-time automatic speech recognition model that transcribes audio into text. Alibaba Cloud lists this snapshot under June 17, 2026; the snapshot ID contains the date June 15,…
Released: June 16, 2026, according to the official announcement. (z.ai) GLM-5.2 is an open-weight flagship model designed for long-horizon agent tasks, particularly software engineering. It accepts…
HappyHorse-1.1-I2V is Alibaba's first-frame image-to-video model, served through Alibaba Cloud Model Studio. It takes one image and an optional text prompt to generate a video with audio. Alibaba…
HappyHorse-1.1-R2V is a reference-image-to-video model offered through Alibaba Cloud Model Studio. It accepts text and up to nine reference images and generates video, with the aim of preserving…
HappyHorse-1.1-T2V is a text-to-video generation model served through Alibaba Cloud Model Studio. Alibaba describes improvements in text understanding, cinematic shot control and motion generation,…
Qwen3.5-OCR is a Qwen3.5-series optical character recognition model offered through Alibaba Cloud Model Studio. It is designed for document parsing, text localization and key-information extraction,…
Qwen-RobotWorld is a language-conditioned video world model for embodied intelligence and part of the Qwen-Robot Suite. Given a current visual observation and a natural-language action instruction, it…
Qwen-RobotManip is a vision-language-action foundation model for robotic manipulation, part of the Qwen-Robot Suite. It combines a Qwen3.5-4B vision-language backbone with a flow-matching Diffusion…
Qwen-RobotNav is a scalable vision-language-navigation model for mobile robots and agentic navigation systems. It is part of the Qwen-Robot Suite and is built on Qwen3-VL. Qwen’s research listing…
Hy-Embodied-0.5-VLA-RoboTwin is a vision-language-action (VLA) checkpoint from Tencent Robotics X and the Tencent Hy Team. It is the supervised fine-tuned RoboTwin variant of Hy-Embodied-0.5-VLA,…