# China AI model release timeline. > Track Chinese AI model release dates and availability across major LLM, image, video, and audio model families. - Canonical page: https://chinaapi.ai/model-updates/ - Human-readable page: https://chinaapi.ai/model-updates/ - Live pricing: https://dash.chinaapi.ai/pricing?lang=en&utm_source=chinaapi&utm_medium=markdown&utm_campaign=model-updates A data-driven wall of Chinese AI model releases. Dates come from the linked vendor sources; ChinaAPI availability is shown on each entry as a secondary status. 3 estimated dates and 5 unknown dates are flagged in the source data. [View live pricing](https://dash.chinaapi.ai/pricing?lang=en&utm_source=chinaapi&utm_medium=model-updates&utm_campaign=pricing) [Browse model pages](https://chinaapi.ai/models/) [Agent ecosystem updates](https://chinaapi.ai/agent-updates/) ## Release wall. 2026-08-03 [Alibaba Cloud Qwen3.8-Max](https://chinaapi.ai/models/qwen3.8-max/) Qwen3.8 Alibaba Cloud released qwen3.8-max on 2026-08-03 as Qwen's most capable flagship model to date. The native vision-language model uses a 2.4T-parameter Mixture-of-Experts architecture and supports reasoning, text generation, and visual understanding, with significant improvements over the Qwen3.7 series. Official Model Studio documentation lists OpenAI-compatible, Anthropic-compatible, and DashScope access across Beijing and multiple global regions. The route is now available through ChinaAPI. LLM ✓ On ChinaAPI confirmed [Source · help.aliyun.com](https://help.aliyun.com/en/model-studio/newly-released-models) 2026-07-31 [MiniMax MiniMax-H3](https://chinaapi.ai/models/minimax-h3/) MiniMax H3 MiniMax released H3 (Hailuo 3.0). The official release notes describe it as a new-generation open general-purpose multimodal video model that understands creative intent across text, image, video, and audio context and delivers more natural, coherent generation. ChinaAPI serves the route for 4-15 second clips at 768P and 2K with native stereo audio in a single pass, billed per second of output and callable through the OpenAI-compatible video endpoint. Video ✓ On ChinaAPI confirmed [Source · platform.minimax.io](https://platform.minimax.io/docs/release-notes/models) 2026-07-31 **ByteDance Doubao-Seedance-2.5** Seedance 2.5 Volcano Engine's Ark model catalog lists Doubao-Seedance-2.5 under model ID doubao-seedance-2-5-260628. Released on 2026-07-31, the video model can generate up to 30 seconds, accept up to 50 reference assets (30 images, 10 videos, and 10 audio clips), and support multimodal reference, video editing, extension, and generation in more than 10 languages. The 260628 suffix remains the model version tag. ChinaAPI has not yet verified this route for customer calls, so no model page or pricing is published yet. Video Not yet confirmed [Source · console.volcengine.com](https://console.volcengine.com/ark/region:cn-beijing/model/detail?name=doubao-seedance-2-5) 2026-07-31 [DeepSeek DeepSeek-V4-Flash-0731](https://chinaapi.ai/models/deepseek-v4-flash/) DeepSeek V4 Flash DeepSeek upgraded its existing deepseek-v4-flash API route with a re-post-trained checkpoint that keeps the preview architecture and model size while strengthening coding and agent capabilities. DeepSeek reports 82.7 on Terminal Bench 2.1, 54.4 on DeepSWE, and 70.3 on Toolathlon Verified using max reasoning effort and an unreleased minimal harness; these are vendor results, not ChinaAPI tests. The official route now natively supports the Responses API and Codex integration, while DeepSeek's web app, mobile app, and V4-Pro API were not part of this update. LLM ✓ On ChinaAPI confirmed [Source · api-docs.deepseek.com](https://api-docs.deepseek.com/updates/) 2026-07-25 **Alibaba Cloud Qwen3.7-Flash** Qwen3.7 Alibaba Cloud released its native vision-language Flash model with 1M context and up to 131K output, upgrading multimodal understanding, real-world and spatial perception, Search Agent and CI Agent execution, and multimodal coding over Qwen3.6-Flash. The official model accepts text, image, and video and exposes built-in tools through the Responses API. ChinaAPI's gateway does not currently list this route. LLM Not yet confirmed [Source · cn.aliyun.com](https://cn.aliyun.com/product/news?category=224&from_alibabacloud=&product=294) 2026-07-16 [Moonshot Kimi K3](https://chinaapi.ai/models/kimi-k3/) Kimi Moonshot released its 2.8T-parameter flagship with 1M context, always-on reasoning, and native image and video understanding. The announced open-source weights are due by July 27, 2026. LLM ✓ On ChinaAPI confirmed [Source · kimi.com](https://www.kimi.com/code/docs/en/kimi-code/whats-new.html) 2026-07-06 [Tencent Hy3](https://chinaapi.ai/models/hunyuan-hy3/) Hunyuan Tencent officially released Hy3, a 256K-context hybrid fast-and-slow-thinking model with improved stability and cost efficiency. LLM ✓ On ChinaAPI confirmed [Source · tencent.com](https://www.tencent.com/tencent-hunyuan-officially-releases-hy3-advancing-agent-capabilities-and-deeper-product-integration/) 2026-06-30 [Meituan LongCat-2.0](https://chinaapi.ai/models/longcat-2.0/) LongCat Meituan released its open MoE reasoning model with 1M context for coding agents and long-horizon tool use. LLM ✓ On ChinaAPI confirmed [Source · meituan.com](https://www.meituan.com/news/NN260630164005904) 2026-06-28 [ByteDance Doubao Seed 2.1 Turbo](https://chinaapi.ai/models/doubao-seed-2-1-turbo-260628/) Doubao Seed Lower-latency Doubao Seed 2.1 model for multimodal reasoning, coding, and production agent workflows. LLM ✓ On ChinaAPI confirmed [Source · volcengine.com](https://www.volcengine.com/product/yunque) 2026-06-28 [ByteDance Doubao Seed 2.1 Pro](https://chinaapi.ai/models/doubao-seed-2-1-pro-260628/) Doubao Seed Flagship Doubao Seed 2.1 model for multimodal reasoning, coding, tool use, and long-horizon agent tasks. LLM ✓ On ChinaAPI confirmed [Source · volcengine.com](https://www.volcengine.com/product/yunque) 2026-06-22 [Alibaba happyhorse-1.1-t2v](https://chinaapi.ai/models/happyhorse-1.1-t2v/) HappyHorse HappyHorse 1.1 text-to-video gateway model with improved semantic understanding and camera control. Video ✓ On ChinaAPI confirmed 2026-06-22 [Alibaba happyhorse-1.1-r2v](https://chinaapi.ai/models/happyhorse-1.1-r2v/) HappyHorse HappyHorse 1.1 reference-to-video gateway model for multi-reference subject, scene, and style consistency. Video ✓ On ChinaAPI confirmed 2026-06-22 [Alibaba happyhorse-1.1-i2v](https://chinaapi.ai/models/happyhorse-1.1-i2v/) HappyHorse HappyHorse 1.1 image-to-video gateway model with stronger texture and identity consistency. Video ✓ On ChinaAPI confirmed 2026-06-16 [Zhipu / Z.ai glm-5.2](https://chinaapi.ai/models/glm-5.2/) GLM Open flagship GLM with 1M context and stronger long-horizon coding and agent workflows. LLM ✓ On ChinaAPI confirmed [Source · huggingface.co](https://huggingface.co/zai-org/GLM-5.2) 2026-06-15 [ByteDance doubao-seedance-2-0-mini-260615](https://chinaapi.ai/models/doubao-seedance-2-0-mini-260615/) Seedance Fast Seedance 2.0 mini video model for lower-cost short-form generation. Video ✓ On ChinaAPI confirmed 2026-06-12 [Moonshot kimi-k2.7-code-highspeed](https://chinaapi.ai/models/kimi-k2.7-code-highspeed/) Kimi High-speed Kimi K2.7 gateway variant for lower-latency coding and agent tasks. LLM ✓ On ChinaAPI confirmed 2026-06-12 [Moonshot kimi-k2.7-code](https://chinaapi.ai/models/kimi-k2.7-code/) Kimi Coding-focused Kimi gateway model for long-horizon repository work with vision support. LLM ✓ On ChinaAPI confirmed 2026-06-02 [Xiaomi mimo-v2.5-asr](https://chinaapi.ai/models/mimo-v2.5-asr/) MiMo Audio MiMo speech recognition for Chinese-English code-switching, Cantonese, Wu, Minnan and Sichuan dialects, noisy and far-field recordings, overlapping speakers, and lyrics. Audio ✓ On ChinaAPI confirmed [Source · mimo.mi.com](https://mimo.mi.com/docs/en-US/updates/model) 2026-06-01 [Alibaba qwen3.7-plus](https://chinaapi.ai/models/qwen3.7-plus/) Qwen Multimodal Qwen gateway model for visual analysis, planning, files, and tool use. LLM ✓ On ChinaAPI confirmed 2026-06-01 [MiniMax MiniMax-M3](https://chinaapi.ai/models/minimax-m3/) MiniMax M Native multimodal M-series model with 1M context, MSA attention, coding, agent, image, and video input support. LLM ✓ On ChinaAPI confirmed [Source · minimax.io](https://www.minimax.io/blog/minimax-m3) 2026-05-28 [StepFun step-3.7-flash](https://chinaapi.ai/models/step-3.7-flash/) Step Fast StepFun gateway model with 256K context, vision input, reasoning, and tool use. LLM ✓ On ChinaAPI confirmed 2026-05-26 **Meituan LongCat-Video-Avatar 1.5** LongCat Video Audio-driven avatar video technical release focused on lip sync, identity consistency, and long-video generation. Video Not yet confirmed [Source · arxiv.org](https://arxiv.org/abs/2605.26486) 2026-05-21 [Alibaba qwen3.7-max](https://chinaapi.ai/models/qwen3.7-max/) Qwen Flagship Qwen gateway model for demanding analysis, coding, and production agent workflows. LLM ✓ On ChinaAPI confirmed 2026-04-27 [Alibaba qwen3.6-flash](https://chinaapi.ai/models/qwen3.6-flash/) Qwen Qwen 3.6 fast vision-language model for visual reasoning, files, and agent workflows. LLM ✓ On ChinaAPI confirmed [Source · en.wikipedia.org](https://en.wikipedia.org/wiki/Qwen) 2026-04-27 [Xiaomi mimo-v2.5-pro](https://chinaapi.ai/models/mimo-v2.5-pro/) MiMo 1.02T-parameter MiMo agent model with 1M context for demanding coding and long-horizon tasks. LLM ✓ On ChinaAPI confirmed [Source · huggingface.co](https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro) 2026-04-27 [Xiaomi mimo-v2.5](https://chinaapi.ai/models/mimo-v2.5/) MiMo Native omnimodal MiMo model with 1M context and text, image, video, and audio understanding. LLM ✓ On ChinaAPI confirmed [Source · huggingface.co](https://huggingface.co/XiaomiMiMo/MiMo-V2.5) 2026-04-25 [Alibaba wan2.7-i2v](https://chinaapi.ai/models/wan2.7-i2v/) Wan Wan 2.7 image-to-video with first/last-frame control, continuation, driving audio, and 720P/1080P output. Video ✓ On ChinaAPI confirmed [Source · help.aliyun.com](https://help.aliyun.com/en/model-studio/model-pricing) 2026-04-24 [DeepSeek deepseek-v4-pro](https://chinaapi.ai/models/deepseek-v4-pro/) DeepSeek V4 Flagship DeepSeek V4 variant with 1.6T total parameters and 1M-token context. LLM ✓ On ChinaAPI confirmed [Source · tomshardware.com](https://www.tomshardware.com/tech-industry/artificial-intelligence/deepseek-launches-1-6-trillion-parameter-v4-on-huawei-chips-as-us-escalates-ai-theft-accusations) 2026-04-24 [DeepSeek deepseek-v4-flash](https://chinaapi.ai/models/deepseek-v4-flash/) DeepSeek V4 Fast DeepSeek V4 variant for efficient chat, coding assistance, and agent loops. LLM ✓ On ChinaAPI confirmed [Source · api-docs.deepseek.com](https://api-docs.deepseek.com/news/news260424/) 2026-04-23 [Xiaomi mimo-v2.5-tts](https://chinaapi.ai/models/mimo-v2.5-tts/) MiMo Audio MiMo expressive text-to-speech with premium built-in voices and natural-language control over pace, emotion, tone, dialect, and character performance. Audio ✓ On ChinaAPI confirmed [Source · mimo.mi.com](https://mimo.mi.com/docs/en-US/updates/model) 2026-04-14 [Moonshot kimi-k2.6](https://chinaapi.ai/models/kimi-k2.6/) Kimi Open native multimodal Kimi model for long-horizon coding, visual design, and agent-swarm workflows. LLM ✓ On ChinaAPI confirmed [Source · huggingface.co](https://huggingface.co/moonshotai/Kimi-K2.6) 2026-04-07 **Zhipu AI GLM-5V-Turbo** GLM A 200K-context multimodal coding foundation model for image, video, file, GUI-agent, and tool-driven workflows. LLM Not yet confirmed [Source · docs.bigmodel.cn](https://docs.bigmodel.cn/cn/update/new-releases) 2026-04-03 [Alibaba wan2.7-videoedit](https://chinaapi.ai/models/wan2.7-videoedit/) Wan Wan 2.7 natural-language video editing model for local/global edits and element replacement. Video ✓ On ChinaAPI confirmed 2026-04-03 [Alibaba wan2.7-t2v](https://chinaapi.ai/models/wan2.7-t2v/) Wan Wan 2.7 text-to-video model for prompt-driven video generation. Video ✓ On ChinaAPI confirmed 2026-04-03 [Alibaba wan2.7-r2v](https://chinaapi.ai/models/wan2.7-r2v/) Wan Wan 2.7 reference-to-video model for controlled generation from visual references. Video ✓ On ChinaAPI confirmed 2026-04-03 [Zhipu / Z.ai glm-5.1](https://chinaapi.ai/models/glm-5.1/) GLM GLM update for longer-horizon tasks, coding agents, and hybrid reasoning. LLM ✓ On ChinaAPI confirmed [Source · huggingface.co](https://huggingface.co/zai-org/GLM-5.1) 2026-04-01 [Alibaba wan2.7-image-pro](https://chinaapi.ai/models/wan2.7-image-pro/) Wan Wan 2.7 Pro image generation and editing with brand-color control, multi-reference consistency, and up to 4096×4096 output. Image ✓ On ChinaAPI confirmed [Source · help.aliyun.com](https://help.aliyun.com/en/model-studio/image-model/) 2026-04-01 [Alibaba wan2.7-image](https://chinaapi.ai/models/wan2.7-image/) Wan Wan 2.7 image model for text-to-image, editing, multi-reference generation, and stronger text rendering. Image ✓ On ChinaAPI confirmed 2026-03 [MiniMax MiniMax-M2.7-highspeed](https://chinaapi.ai/models/minimax-m2.7-highspeed/) MiniMax M High-speed MiniMax variant for low-latency coding and agent workflows. LLM ✓ On ChinaAPI estimated [Source · minimax.io](https://www.minimax.io/) 2026-03 [MiniMax MiniMax-M2.7](https://chinaapi.ai/models/minimax-m2.7/) MiniMax M Open-weight MiniMax model for chat, coding, office work, and agentic tasks. LLM ✓ On ChinaAPI estimated [Source · minimax.io](https://www.minimax.io/) 2026-02-16 **Alibaba Qwen3.5** Qwen Native multimodal-agent Qwen release that preceded the served Qwen 3.6 and 3.7 gateway models. LLM Not yet confirmed [Source · en.wikipedia.org](https://en.wikipedia.org/wiki/Qwen) 2026-02-11 [Zhipu / Z.ai glm-5-turbo](https://chinaapi.ai/models/glm-5-turbo/) GLM Efficient GLM-5 turbo variant for faster reasoning, coding, and agent work. LLM ✓ On ChinaAPI confirmed [Source · huggingface.co](https://huggingface.co/zai-org/GLM-5) 2026-02-11 [Zhipu / Z.ai glm-5](https://chinaapi.ai/models/glm-5/) GLM GLM-5 release for coding, reasoning, and agent workflows. LLM ✓ On ChinaAPI confirmed [Source · huggingface.co](https://huggingface.co/zai-org/GLM-5) 2026-02-05 [Kuaishou kling-v3-omni](https://chinaapi.ai/models/kling-v3-omni/) Kling Kling 3.0 Omni variant for reference-to-video and editing workflows. Video ✓ On ChinaAPI confirmed [Source · en.wikipedia.org](https://en.wikipedia.org/wiki/Kling_AI) 2026-02-05 [Kuaishou kling-v3](https://chinaapi.ai/models/kling-v3/) Kling Kling 3.0 generation with text-to-video, image-to-video, audio, and 720p to 4K output tiers. Video ✓ On ChinaAPI confirmed [Source · en.wikipedia.org](https://en.wikipedia.org/wiki/Kling_AI) 2026-02-05 [Kuaishou kling-3.0-turbo](https://chinaapi.ai/models/kling-3.0-turbo/) Kling Fast Kling 3.0 Turbo path for lower-latency video generation with audio. Video ✓ On ChinaAPI confirmed [Source · en.wikipedia.org](https://en.wikipedia.org/wiki/Kling_AI) 2026-02-01 **StepFun Step-3.5-Flash** Step Open StepFun flash model with 256K context and tool-use support. LLM Not yet confirmed [Source · huggingface.co](https://huggingface.co/stepfun-ai/Step-3.5-Flash) 2026-01-28 [ByteDance doubao-seedream-5-0-260128](https://chinaapi.ai/models/doubao-seedream-5-0-260128/) Seedream Seedream 5.0-lite image model with controllable creation, retrieval, and stronger subject consistency. Image ✓ On ChinaAPI confirmed 2026-01-28 [ByteDance doubao-seedance-2-0-fast-260128](https://chinaapi.ai/models/doubao-seedance-2-0-fast-260128/) Seedance Fast Seedance 2.0 video generation option for cost-efficient clips. Video ✓ On ChinaAPI confirmed 2026-01-28 [ByteDance doubao-seedance-2-0-260128](https://chinaapi.ai/models/doubao-seedance-2-0-260128/) Seedance Seedance 2.0 flagship video generation model billed by output resolution. Video ✓ On ChinaAPI confirmed 2026-01-23 [MiniMax Speech 2.8 Turbo](https://chinaapi.ai/models/speech-2.8-turbo/) Speech 2.8 MiniMax released Speech 2.8 with native sound tags, improved voice cloning, studio-grade audio, and cross-lingual speech; Turbo is its speed-focused option. Audio ✓ On ChinaAPI confirmed [Source · minimax.io](https://www.minimax.io/news/minimax-speech-28) 2026-01-23 [MiniMax Speech 2.8 HD](https://chinaapi.ai/models/speech-2.8-hd/) Speech 2.8 MiniMax released Speech 2.8 with native sound tags, improved voice cloning, studio-grade audio, and cross-lingual speech; HD is its high-fidelity option. Audio ✓ On ChinaAPI confirmed [Source · minimax.io](https://www.minimax.io/news/minimax-speech-28) 2026-01-06 [Alibaba Qwen3-ASR-Flash adds OpenAI-compatible mode](https://chinaapi.ai/models/qwen3-asr-flash/) Qwen Audio Alibaba added OpenAI-compatible mode to Qwen3-ASR-Flash, retaining multilingual transcription, ITN, and streaming support. Audio ✓ On ChinaAPI confirmed [Source · help.aliyun.com](https://help.aliyun.com/en/model-studio/newly-released-models) 2026-01-01 **Moonshot kimi-k2.5** Kimi Open native multimodal Kimi agent model with 256K context and vision-language training. LLM Not yet confirmed [Source · huggingface.co](https://huggingface.co/moonshotai/Kimi-K2.5) 2025-12-10 **Zhipu AI GLM-ASR-2512** GLM Audio Z.AI released its multilingual ASR model with improved custom-dictionary support and specialised-terminology recognition. Audio Not yet confirmed [Source · docs.z.ai](https://docs.z.ai/release-notes/new-released) 2025-12 **Meituan LongCat-Image** LongCat Small image-generation model focused on data quality over model size. Image Not yet confirmed [Source · en.wikipedia.org](https://en.wikipedia.org/wiki/Meituan) 2025-10-28 [MiniMax MiniMax-Hailuo-2.3-Fast](https://chinaapi.ai/models/minimax-hailuo-2.3-fast/) Hailuo Fast Hailuo 2.3 path for lower-cost image-to-video generation. Video ✓ On ChinaAPI confirmed [Source · minimax.io](https://www.minimax.io/news/minimax-hailuo-23) 2025-10-28 [MiniMax MiniMax-Hailuo-2.3](https://chinaapi.ai/models/minimax-hailuo-2.3/) Hailuo Hailuo 2.3 improves complex motion, physical actions, stylization, and micro-expression performance. Video ✓ On ChinaAPI confirmed [Source · minimax.io](https://www.minimax.io/news/minimax-hailuo-23) 2025-09-22 [Alibaba Qwen3-TTS-Flash](https://chinaapi.ai/models/qwen3-tts-flash/) Qwen Audio Alibaba released Qwen3-TTS-Flash for offline speech synthesis with expressive voices, multilingual and dialect output, and low-latency API access. Audio ✓ On ChinaAPI confirmed [Source · help.aliyun.com](https://help.aliyun.com/en/model-studio/newly-released-models) 2025-06 [MiniMax MiniMax-Hailuo-02](https://chinaapi.ai/models/minimax-hailuo-02/) Hailuo Budget Hailuo 02 video model for text-to-video and image-to-video generation from 512p. Video ✓ On ChinaAPI estimated [Source · minimax.io](https://www.minimax.io/news/minimax-hailuo-23) Release date not publicly confirmed [StepFun StepAudio 2.5 TTS](https://chinaapi.ai/models/stepaudio-2.5-tts/) StepAudio Context-aware text-to-speech with global and inline natural-language direction, expressive delivery, and direct OpenAI-compatible /v1/audio/speech output. Audio ✓ On ChinaAPI unknown [Source · platform.stepfun.com](https://platform.stepfun.com/docs/zh/guides/models/stepaudio-2.5-tts) Release date not publicly confirmed [StepFun StepAudio 2.5 ASR](https://chinaapi.ai/models/stepaudio-2.5-asr/) StepAudio Fast 4B MTP streaming transcription for Chinese-English audio with ITN normalization and an OpenAI-compatible transcription route. Audio ✓ On ChinaAPI unknown [Source · platform.stepfun.com](https://platform.stepfun.com/docs/zh/guides/models/audio) Release date not publicly confirmed [StepFun Step TTS Mini](https://chinaapi.ai/models/step-tts-mini/) Step TTS Cost-efficient multilingual speech synthesis with official and cloned voices, pronunciation mapping, and style controls. Audio ✓ On ChinaAPI unknown [Source · platform.stepfun.com](https://platform.stepfun.com/docs/zh/guides/models/overview) Release date not publicly confirmed [StepFun Step TTS 2](https://chinaapi.ai/models/step-tts-2/) Step TTS Expressive speech synthesis with official and cloned voices, pronunciation mapping, style controls, and OpenAI-compatible output. Audio ✓ On ChinaAPI unknown [Source · platform.stepfun.com](https://platform.stepfun.com/docs/zh/api-reference/audio/create-audio) Release date not publicly confirmed [Zhipu AI GLM-TTS](https://chinaapi.ai/models/glm-tts/) GLM Audio Context-aware expressive speech synthesis with streaming output and controls for voice, speed, and volume. Audio ✓ On ChinaAPI unknown [Source · docs.bigmodel.cn](https://docs.bigmodel.cn/cn/guide/models/sound-and-video/glm-tts) --- Markdown twin of https://chinaapi.ai/model-updates/ — generated from its canonical HTML by scripts/gen_markdown_twins.py. Full site reference: https://chinaapi.ai/llms-full.txt