DeepSeek-V3.1-Base
DeepSeek-V3.1-Base is the base pretrained language model for DeepSeek-V3.1. It was further trained for long-context extension on the original DeepSeek-V3 base model checkpoint; the post-trained…
MODEL DIRECTORY
New releases, in one place. Explore models by their publication date.
DeepSeek-V3.1-Base is the base pretrained language model for DeepSeek-V3.1. It was further trained for long-context extension on the original DeepSeek-V3 base model checkpoint; the post-trained…
DeepSeek-V3.1 was released on August 21, 2025. It is a hybrid reasoning model for dialogue, reasoning, and agent tasks, with both thinking and non-thinking modes supported by the same model. At…
Release date: August 11, 2025. (huggingface.co) GLM-4.5V is a vision-language model for multimodal understanding and visual reasoning. Based on GLM-4.5-Air, it uses a mixture-of-experts (MoE)…
Wan2.2-I2V-Flash is a fast image-to-video model in Tongyi Wanxiang 2.2, with inference services provided by Alibaba Cloud Bailian. The date listed in the official release record is 2025-08-11; the…
GPT-5 pro is a deep reasoning version released by OpenAI alongside GPT-5 on August 7, 2025, for highly difficult and complex tasks. The official system card calls this configuration in ChatGPT at…
In the GPT-5 System Card, OpenAI identifies gpt-5-main-mini as a fast, high-throughput model and lists it as a successor to GPT-4o mini. It is a different model from gpt-5-thinking-mini, the deep…
GPT-5 Chat is a non-reasoning model in the GPT-5 series that was previously used in ChatGPT, distinct from the gpt-5 reasoning model in the API. The official model page states that this ID points to…
GPT-5 nano is a lightweight model in the GPT-5 family for speed- and cost-sensitive tasks. Its API model ID is gpt-5-nano, and its pinned version snapshot is gpt-5-nano-2025-08-07. (developers.openai.…
GPT-5 mini is a faster, lower-cost version of GPT-5, suited to cost-sensitive, low-latency, and high-throughput scenarios, particularly workloads with well-defined tasks and precise prompts.…
GPT-5 is a reasoning model released by OpenAI for developers on August 7, 2025, with a focus on programming and agentic tasks requiring multiple steps. gpt-5-2025-08-07 is its pinned version snapshot;…
MiniMax Speech 2.5 is a speech generation model for multilingual text-to-speech and voice cloning, and a successor to Speech 02. It was officially released and announced for global availability on…
Qianfan-Correct is Baidu's domain-specialized model for writing instruction, focused on everyday essay grading. Baidu officially classifies it as a text generation model in the Qianfan Inference…
An introduction in this language is coming soon.
ERNIE 4.5 Turbo VL is Baidu's vision-language model for image understanding. Baidu's official model update log states that its image understanding, content creation, translation, and coding…
Claude Opus 4.1 is Anthropic's upgraded version of Claude Opus 4, focusing on improvements in agentic tasks, real-world programming, and reasoning. It was released on August 5, 2025. (anthropic.com)…
Genie 3 is a general-purpose world model developed by Google DeepMind for generating interactive, dynamic environments. The Genie 3 research preview was officially announced on August 5, 2025,…
Gemini 2.5 Deep Think is an enhanced reasoning mode for Gemini 2.5 Pro, designed for complex mathematics, programming, and problems requiring creative solutions. Google announced the rollout of this…
Qwen-Flash is a text generation and reasoning model in the Qwen3 series. The official international site lists qwen-flash and the snapshot version qwen-flash-2025-07-28 in its update of August 1, 2025…