AI Model Releases
New models from 26 AI labs, straight from their official announcements, blogs and open-weights repositories. Updated every 30 minutes · last checked 2026-10-01 14:30 UTC.
760 releases · page 10 / 16
| Company | Model | Released at (UTC) |
|---|---|---|
| StepFun | StepFun-Prover-Preview-7B stepfun-ai/StepFun-Prover-Preview-7BText generation model released as open weights by StepFun on Hugging Face (library: transformers, license: apache-2.0). | |
| StepFun | NextStep-1-Large stepfun-ai/NextStep-1-LargeText-to-image model released as open weights by StepFun on Hugging Face (library: transformers, license: apache-2.0). | |
| Cohere | Command-a-reasoning-08-2025 CohereLabs/command-a-reasoning-08-2025Text generation model released as open weights by Cohere on Hugging Face (library: transformers, license: cc-by-nc-4.0). | |
| ByteDance Seed | M3-Agent-Memorization ByteDance-Seed/M3-Agent-MemorizationModel released as open weights by ByteDance Seed on Hugging Face (license: apache-2.0). | |
| ByteDance Seed | M3-Agent-Control ByteDance-Seed/M3-Agent-ControlModel released as open weights by ByteDance Seed on Hugging Face (license: apache-2.0). | |
| StepFun | NextStep-1-Large-Edit stepfun-ai/NextStep-1-Large-EditImage editing model released as open weights by StepFun on Hugging Face (library: transformers, license: apache-2.0). | |
| Zhipu AI (GLM) | GLM-4.5V zai-org/GLM-4.5VVision-language model released as open weights by Zhipu AI (GLM) on Hugging Face (library: transformers, license: mit). | |
| Stability AI | stable-diffusion-3.5-controlnets-tensorrt stabilityai/stable-diffusion-3.5-controlnets-tensorrtText-to-image model released as open weights by Stability AI on Hugging Face (library: tensorrt, license: other). | |
| Stability AI | stable-diffusion-3.5-medium-tensorrt stabilityai/stable-diffusion-3.5-medium-tensorrtText-to-image model released as open weights by Stability AI on Hugging Face (library: tensorrt, license: other). | |
| OpenAI | GPT-5 Introducing GPT-5 for developersIntroducing GPT-5 in our API platform—offering high reasoning performance, new controls for devs, and best-in-class results on real coding tasks. | |
| Xiaomi MiMo | MiMo-VL-7B-RL-2508 XiaomiMiMo/MiMo-VL-7B-RL-2508Vision-language model released as open weights by Xiaomi MiMo on Hugging Face (library: transformers, license: mit). | |
| Xiaomi MiMo | MiMo-VL-7B-SFT-2508 XiaomiMiMo/MiMo-VL-7B-SFT-2508Vision-language model released as open weights by Xiaomi MiMo on Hugging Face (library: transformers, license: mit). | |
| OpenAI | GPT-5 Introducing GPT-5We are introducing GPT‑5, our best AI system yet. GPT‑5 is a significant leap in intelligence over all our previous models, featuring state-of-the-art performance across coding, math, writing, health, visual perception, and more. | |
| OpenAI | GPT-5 GPT-5 System CardThis GPT-5 system card explains how a unified model routing system powers fast and smart responses using gpt-5-main, gpt-5-thinking, and lightweight versions like gpt-5-thinking-nano, optimized for different tasks and developer use. | |
| Anthropic | Claude Opus 4.1 Claude Opus 4.1Our most powerful model for handling complex agent and coding tasks | |
| OpenAI | gpt-oss Introducing gpt-ossWe’re releasing gpt-oss-120b and gpt-oss-20b—two state-of-the-art open-weight language models that deliver strong real-world performance at low cost. Available under the flexible Apache 2.0 license, these models outperform similarly sized open models on reasoning tasks… | |
| OpenAI | gpt-oss-120b gpt-oss-120b & gpt-oss-20b Model CardWe introduce gpt-oss-120b and gpt-oss-20b, two open-weight reasoning models available under the Apache 2.0 license and our gpt-oss usage policy. | |
| OpenAI | gpt-oss-20b openai/gpt-oss-20bText generation model released as open weights by OpenAI on Hugging Face (library: transformers, license: apache-2.0). | |
| OpenAI | gpt-oss-120b openai/gpt-oss-120bText generation model released as open weights by OpenAI on Hugging Face (library: transformers, license: apache-2.0). | |
| Alibaba Qwen | Qwen-Image Qwen-Image: Crafting with Native Text RenderingGITHUB HUGGING FACE MODELSCOPE DEMO DISCORD We are thrilled to release Qwen-Image, a 20B MMDiT image foundation model that achieves significant advances in complex text rendering and precise image editing. To try the latest model, feel free to visit Qwen Chat and choose “Image… | |
| ByteDance Seed | cudaLLM-8B ByteDance-Seed/cudaLLM-8BText generation model released as open weights by ByteDance Seed on Hugging Face (license: apache-2.0). | |
| Mistral AI | Codestral 25.08 Announcing Codestral 25.08 and the Complete Mistral Coding Stack for Enterprise | |
| StepFun | Step3 stepfun-ai/step3Vision-language model released as open weights by StepFun on Hugging Face (library: transformers, license: apache-2.0). | |
| Cohere | Command-a-vision-07-2025 CohereLabs/command-a-vision-07-2025Vision-language model released as open weights by Cohere on Hugging Face (library: transformers, license: cc-by-nc-4.0). | |
| Alibaba Qwen | Wan2.2-T2V-A14B Wan-AI/Wan2.2-T2V-A14BText-to-video model released as open weights by Alibaba Qwen on Hugging Face (library: wan2.2, license: apache-2.0). | |
| Alibaba Qwen | Qwen-MT Qwen-MT: Where Speed Meets Smart TranslationDEMO API DISCORD Introduction Here we introduce the latest update of Qwen-MT (qwen-mt-turbo) via Qwen API. This update builds upon the powerful Qwen3, leveraging trillions multilingual and translation tokens to comprehensively enhance the model’s multilingual understanding and… | |
| Alibaba Qwen | Wan2.2-I2V-A14B Wan-AI/Wan2.2-I2V-A14BImage-to-video model released as open weights by Alibaba Qwen on Hugging Face (library: wan2.2, license: apache-2.0). | |
| Alibaba Qwen | Qwen3-Coder Qwen3-Coder: Agentic Coding in the WorldGITHUB HUGGING FACE MODELSCOPE DISCORD Today, we’re announcing Qwen3-Coder, our most agentic code model to date. Qwen3-Coder is available in multiple sizes, but we’re excited to introduce its most powerful variant first: Qwen3-Coder-480B-A35B-Instruct — a… | |
| Kuaishou (Kling / Kolors) | KAT-V1-40B Kwaipilot/KAT-V1-40BText generation model released as open weights by Kuaishou (Kling / Kolors) on Hugging Face (library: transformers, license: other). | |
| Zhipu AI (GLM) | GLM-4.5-Base zai-org/GLM-4.5-BaseText generation model released as open weights by Zhipu AI (GLM) on Hugging Face (library: transformers, license: mit). | |
| Zhipu AI (GLM) | GLM-4.5-Air zai-org/GLM-4.5-AirText generation model released as open weights by Zhipu AI (GLM) on Hugging Face (library: transformers, license: mit). | |
| Zhipu AI (GLM) | GLM-4.5 zai-org/GLM-4.5Text generation model released as open weights by Zhipu AI (GLM) on Hugging Face (library: transformers, license: mit). | |
| Zhipu AI (GLM) | GLM-4.5-Air-Base zai-org/GLM-4.5-Air-BaseText generation model released as open weights by Zhipu AI (GLM) on Hugging Face (library: transformers, license: mit). | |
| Alibaba Qwen | Wan2.2-TI2V-5B Wan-AI/Wan2.2-TI2V-5BText-to-video model released as open weights by Alibaba Qwen on Hugging Face (library: wan2.2, license: apache-2.0). | |
| Mistral AI | Magistral-Small-2507 mistralai/Magistral-Small-2507Model released as open weights by Mistral AI on Hugging Face (library: vllm, license: apache-2.0). | |
| ByteDance Seed | Seed-X-RM-7B ByteDance-Seed/Seed-X-RM-7BTranslation model released as open weights by ByteDance Seed on Hugging Face (license: other). | |
| ByteDance Seed | Seed-X-Instruct-7B ByteDance-Seed/Seed-X-Instruct-7BTranslation model released as open weights by ByteDance Seed on Hugging Face (license: other). | |
| ByteDance Seed | Seed-X-PPO-7B ByteDance-Seed/Seed-X-PPO-7BTranslation model released as open weights by ByteDance Seed on Hugging Face (license: other). | |
| Mistral AI | Voxtral Voxtral | |
| Moonshot AI (Kimi) | Kimi-K2-Instruct moonshotai/Kimi-K2-InstructText generation model released as open weights by Moonshot AI (Kimi) on Hugging Face (library: transformers, license: other). | |
| Mistral AI | Devstral Upgrading agentic coding capabilities with the new Devstral models | |
| xAI | Grok 4 Grok 4Grok 4 is the most intelligent model in the world. It includes native tool use and real-time search integration, and is available now to SuperGrok and Premium+ subscribers, as well as through the xAI API. We are also introducing a new SuperGrok Heavy tier with access to Grok 4… | |
| Baidu (ERNIE) | ERNIE-4.5-300B-A47B-2Bits-TP4-Paddle baidu/ERNIE-4.5-300B-A47B-2Bits-TP4-PaddleText generation model released as open weights by Baidu (ERNIE) on Hugging Face (license: apache-2.0). | |
| Baidu (ERNIE) | ERNIE-4.5-300B-A47B-2Bits-TP2-Paddle baidu/ERNIE-4.5-300B-A47B-2Bits-TP2-PaddleText generation model released as open weights by Baidu (ERNIE) on Hugging Face (license: apache-2.0). | |
| Black Forest Labs | FLUX.1-Krea-dev black-forest-labs/FLUX.1-Krea-devText-to-image model released as open weights by Black Forest Labs on Hugging Face (library: diffusers, license: other). | |
| Moonshot AI (Kimi) | Kimi-K2-Base moonshotai/Kimi-K2-BaseText generation model released as open weights by Moonshot AI (Kimi) on Hugging Face (library: transformers, license: other). | |
| AI21 Labs | AI21-Jamba-Large-1.7 ai21labs/AI21-Jamba-Large-1.7Model released as open weights by AI21 Labs on Hugging Face (license: other). | |
| ByteDance Seed | VINCIE-3B ByteDance-Seed/VINCIE-3BImage editing model released as open weights by ByteDance Seed on Hugging Face (library: vincie, license: apache-2.0). | |
| ByteDance Seed | Tar-TA-Tok ByteDance-Seed/Tar-TA-TokModel released as open weights by ByteDance Seed on Hugging Face (license: apache-2.0). | |
| ByteDance Seed | Tar-1.5B ByteDance-Seed/Tar-1.5BOmni (any-to-any) model released as open weights by ByteDance Seed on Hugging Face (license: apache-2.0). |