{"title":"Model Release Velocity","description":"Release frequency, developer share, model availability, and cadence from verified model launches.","dataset_id":"spd:release-velocity","canonical_url":"https://superpowerdaily.com/research/release-velocity","version_url":"https://superpowerdaily.com/research/release-velocity/versions/v7","version":"v7","snapshot_hash":"e7fb6b6162760400557eb0124b9c63722247fe8c38df5f7676e189360138078a","date_created":"2026-08-27T01:49:17.056Z","date_modified":"2026-08-27T16:10:58.284Z","license":{"name":"Superpower Daily data reuse terms","url":"https://superpowerdaily.com/terms"},"license_url":"https://superpowerdaily.com/terms","temporal_coverage":"2024-10/2026-08-27","coverage_note":"Release counts reflect the monitored source set and verified records, not every model published globally.","measurement_technique":["Each verified model-launch record contributes one release observation.","Developer names are normalized during extraction; series preserve currently disclosed attribution.","Release volume measures activity, not capability or adoption."],"methodology":["Each verified model-launch record contributes one release observation.","Developer names are normalized during extraction; series preserve currently disclosed attribution.","Release volume measures activity, not capability or adoption."],"metrics":[{"label":"Verified releases","value":"34","detail":"34 total model launches"},{"label":"Developers","value":"24","detail":"Named release organizations"},{"label":"Most active","value":"OpenAI","detail":"6 releases"},{"label":"Open share","value":"32%","detail":"Disclosed open availability/license"}],"columns":[{"key":"date","label":"Date"},{"key":"model","label":"Model"},{"key":"developer","label":"Developer"},{"key":"release_type","label":"Release type"},{"key":"availability","label":"Availability"},{"key":"license","label":"License"}],"data":[{"date":"2024-10-07","model":"Inflection 3.0","license":null,"developer":"Inflection AI","availability":"Commercial API available to developers; enterprise deployment on-premises, in the cloud, or hybrid","release_type":"commercial API and enterprise AI system"},{"date":"2025-08-22","model":"GPT-4b micro","license":null,"developer":"OpenAI","availability":"developed for research purposes; not broadly available","release_type":"experimental research"},{"date":"2025-12-11","model":"GPT-5.2","license":null,"developer":"OpenAI","availability":"In ChatGPT: GPT-5.2 Instant, Thinking, and Pro begin rolling out today to paid plans; In the API: available now to all developers.","release_type":"API and ChatGPT (model variants: GPT-5.2 Instant, GPT-5.2 Thinking, GPT-5.2 Pro)"},{"date":"2025-12-16","model":"GPT Image 1.5","license":null,"developer":"OpenAI","availability":"Model rolling out in ChatGPT for all users and available in the API as GPT Image 1.5; the new Images experience is rolling out to most ChatGPT users, with Business and Enterprise access coming later; global.","release_type":"API and ChatGPT rollout"},{"date":"2026-04-07","model":"Claude Mythos Preview","license":null,"developer":"Anthropic","availability":"restricted access (restricted-access rollout)","release_type":"Preview"},{"date":"2026-04-28","model":"Laguna XS.2","license":"Apache 2.0","developer":"Poolside","availability":"Preview; Hugging Face, Poolside API, and OpenRouter","release_type":"open weights"},{"date":"2026-06-09","model":"Claude Fable 5","license":null,"developer":"Anthropic","availability":"Publicly accessible via API and Enterprise; included in subscription plans through 2026-06-22 after which access requires purchase of usage credits; some sensitive-topic access restricted and Mythos 5 limited to vetted Project Glasswing participants","release_type":"Public release (Mythos-class); available via API and Enterprise"},{"date":"2026-06-30","model":"Claude Sonnet 5","license":null,"developer":"Anthropic","availability":"All Claude plans and Claude API","release_type":"general availability and API"},{"date":"2026-07-02","model":"Laguna XS 2.1","license":"OpenMDW-1.1","developer":"Poolside","availability":"Poolside API, OpenRouter, Hugging Face weights, and local deployment","release_type":"open weights"},{"date":"2026-07-08","model":"GPT-Live","license":null,"developer":"OpenAI","availability":"Rolling out to ChatGPT users globally (ChatGPT Voice); API availability planned/coming soon. (Update 2026-07-31: SynthID watermarking and API verification access added for supported audio.)","release_type":"voice model (full-duplex)"},{"date":"2026-07-09","model":"GPT-5.6 (Sol, Terra, Luna)","license":null,"developer":"OpenAI","availability":"General availability following a limited preview","release_type":"general availability"},{"date":"2026-07-16","model":"Kimi K3","license":"License requires companies operating a model-as-a-service business with > $20M total annual revenue (over 12 months) to enter a separate agreement with Moonshot before commercial use.","developer":"Moonshot","availability":"Weights published for anyone to download; model can be run locally on user hardware","release_type":"open-weights release (weights published for public download)"},{"date":"2026-07-24","model":"Claude Opus 5","license":null,"developer":"Anthropic","availability":"Available immediately; default on Claude Max and strongest model on Claude Pro","release_type":"general availability"},{"date":"2026-07-31","model":"MiniMax H3","license":"MiniMax H3 Community License Agreement","developer":"MiniMax","availability":"Hosted API globally; downloadable weights restricted to the license's applicable territories","release_type":"API and open weights"},{"date":"2026-08-10","model":"Muse Glimmer","license":"Apache 2.0","developer":"Meta","availability":"Consumer hardware with 24 GB or 32 GB of memory","release_type":"open weights"},{"date":"2026-08-11","model":"Nemotron 3.5 Lightning","license":"OpenMDW-1.1","developer":"Nvidia","availability":"Available through Nvidia and model repositories including Hugging Face and ModelScope; NeMo Switchyard available on GitHub","release_type":"open weights (public release)"},{"date":"2026-08-12","model":"Qwen3.8-27B-CRACK-GGUF","license":null,"developer":"Jinho Jang (dealign.ai)","availability":"Hugging Face repository; downloadable GGUF files and a 0.9 GB vision projector for local use","release_type":"Open weights; GGUF quantizations packaged for local inference via llama.cpp"},{"date":"2026-08-16","model":"GLM-5.3","license":null,"developer":"Zhipu (Z.ai)","availability":null,"release_type":null},{"date":"2026-08-18","model":"GPT-5.6 Sol","license":null,"developer":"OpenAI","availability":"Limited release to a small group of customers; wider access planned later","release_type":"API"},{"date":"2026-08-19","model":"Ornith-1.5","license":null,"developer":"Ornith AI","availability":"Hugging Face collection; local deployment and quantized mobile version described","release_type":"open weights"},{"date":"2026-08-21","model":"deepseek-v4-flash-vision-exp","license":null,"developer":"DeepSeek","availability":"Available through Chat Completions, Messages, and Responses requests","release_type":"experimental API model"},{"date":"2026-08-21","model":"Falcon 2","license":null,"developer":"Murf AI","availability":null,"release_type":"proprietary foundation voice model"},{"date":"2026-08-21","model":"V4-Flash-Vision-Exp","license":null,"developer":"DeepSeek","availability":"DeepSeek API","release_type":"experimental API model"},{"date":"2026-08-24","model":"HiDream-O1-World","license":null,"developer":"HiDream.ai","availability":null,"release_type":"model launch"},{"date":"2026-08-24","model":"Thomson","license":"Non-commercial license for the planned smaller open-weight version","developer":"Thomson Reuters","availability":"Initially available in CoCounsel Legal’s Tabular Analysis feature; smaller version planned for Hugging Face","release_type":"enterprise launch; open-weight version planned"},{"date":"2026-08-24","model":"GEN-1.5","license":null,"developer":"Generalist AI","availability":"Available through direct partnerships on Generalist AI's own robot fleet; no public weights, API, or self-serve product","release_type":"research release"},{"date":"2026-08-26","model":"Qwen3.8-Flash-Next","license":null,"developer":"Alibaba Qwen team","availability":"Weights on Hugging Face and ModelScope; production Qwen3.8-Flash through QwenCloud; API expected shortly","release_type":"architecture preview; open weights; production API"},{"date":"2026-08-26","model":"Granite 4.2","license":"Apache 2.0","developer":"IBM","availability":"Hugging Face, Ollama, GitHub, and other platforms","release_type":"open weights"},{"date":"2026-08-26","model":"Grok 4.6","license":null,"developer":"xAI","availability":"Microsoft Foundry for Azure customers","release_type":"public preview"},{"date":"2026-08-26","model":"Qwen3.8-Flash","license":null,"developer":"Alibaba","availability":"API access","release_type":"API"},{"date":"2026-08-26","model":"Qwen3.8-Flash","license":null,"developer":"Alibaba Qwen","availability":"QwenCloud managed API","release_type":"API"},{"date":"2026-08-27","model":"GLM-5.3-Flash","license":"MIT","developer":"Z.ai","availability":"Z.ai API; weights available on Hugging Face","release_type":"API and open weights"},{"date":"2026-08-27","model":"Gemini 3.5 Transcribe","license":null,"developer":"Google","availability":"Google AI Studio; Gemini Enterprise Agent Platform; Gboard for Android; Gemini app on macOS; Chrome support coming soon","release_type":"API"},{"date":"2026-08-27","model":"leiolai-1","license":null,"developer":"Leiolai","availability":"Consumer app and developer API","release_type":"consumer app and developer API"}],"sources":[{"url":"https://openai.com/index/gpt-5-6","name":"OpenAI News","title":"GPT-5.6: Frontier intelligence that scales with your ambition","records":1},{"url":"https://siliconangle.com/2026/08/24/thomson-reuters-launches-proprietary-ai-model-for-legal-work/","name":"Google News AI","title":"Thomson Reuters launches proprietary AI model for legal work - SiliconANGLE","records":1},{"url":"https://thelogic.co/news/thomson-reuters-custom-ai-launch/","name":"Google News AI Model Launch Tracker","title":"Thomson Reuters launches its own AI model to reduce reliance on big tech - The Logic","records":1},{"url":"https://www.stocktitan.net/news/TRI/thomson-reuters-leverages-its-world-class-data-assets-to-launch-its-6yterby5xq24.html","name":"Google News AI Model Launch Tracker","title":"Thomson Reuters launches its own AI model for legal document reviews","records":1},{"url":"https://the-decoder.com/thomson-reuters-bets-40m-on-owning-its-ai-instead-of-renting-from-openai-or-anthropic/","name":"The Decoder","title":"Thomson Reuters bets $40M on owning its AI instead of renting from OpenAI or Anthropic","records":1},{"url":"https://www.marktechpost.com/2026/08/24/generalist-ai-releases-gen-1-5-a-robot-foundation-model-that-learns-new-tasks-from-one-3-12-second-demo/","name":"MarkTechPost","title":"Generalist AI Releases GEN-1.5: A Robot Foundation Model That Learns New Tasks From One 3–12 Second Demo","records":1},{"url":"https://vir.com.vn/hidreamai-releases-hidream-o1-world-interactive-artificial-intelligence-model-159321.html","name":"Google News AI Model Launch Tracker","title":"HiDream.ai releases HiDream-O1-World interactive artificial intelligence model","records":1},{"url":"https://www.readthepeak.com/p/thomson-reuters-launches-its-own-ai-model","name":"Google News AI Model Launch Tracker","title":"Thomson Reuters Launches Its Own AI Model","records":1},{"url":"https://the-decoder.com/ibm-drops-open-weight-granite-4-2-family-with-built-in-agentic-capabilities-under-apache-2-0/","name":"The Decoder","title":"IBM drops open-weight Granite 4.2 family with built-in agentic capabilities under Apache 2.0","records":1},{"url":"https://arstechnica.com/ai/2026/08/ibms-new-granite-4-2-models-ride-the-wave-of-interest-in-local-llms/","name":"Ars Technica AI","title":"IBM's new Granite 4.2 models ride the wave of interest in local LLMs","records":1},{"url":"https://www.yahoo.com/news/articles/alibabas-qwen-launches-qwen3-8-150149621.html","name":"Google News AI Model Launch Tracker","title":"Alibaba's Qwen launches Qwen3.8-Flash AI model with lower training costs","records":1},{"url":"https://m.economictimes.com/tech/artificial-intelligence/alibabas-qwen-launches-qwen3-8-flash-ai-model-with-lower-training-costs/articleshow/133544818.cms","name":"Google News AI Model Launch Tracker","title":"Alibaba's Qwen launches Qwen3.8-Flash AI model with lower training costs - The Economic Times","records":1},{"url":"https://the-decoder.com/alibaba-releases-qwen3-8-flash-next-targeting-ultimate-cost-efficiency/","name":"The Decoder","title":"Alibaba releases Qwen3.8-Flash-Next, targeting \"ultimate cost efficiency\"","records":1},{"url":"https://www.marktechpost.com/2026/08/26/alibabas-qwen-team-releases-qwen3-8-flash-next-a-125b-multimodal-moe-with-6b-active-parameters-previewing-the-qwen4-architecture/","name":"MarkTechPost","title":"Alibaba's Qwen Team Releases Qwen3.8-Flash-Next: A 125B Multimodal MoE With 6B Active Parameters Previewing the Qwen4 Architecture","records":1},{"url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-3-5-transcribe/","name":"Hacker News AI","title":"Intelligent transcription with Gemini 3.5 Transcribe","records":1},{"url":"https://cryptobriefing.com/zhipu-glm-53-flash-chinese-ai-chips/","name":"Google News AI Model Launch Tracker","title":"Zhipu launches GLM-5.3-Flash, its first natively multimodal model built for Chinese chips","records":1},{"url":"https://autoclaw.z.ai/models/","name":"Z.ai Official","title":"AutoClaw Models | GLM Model Support","records":1},{"url":"https://arstechnica.com/ai/2026/08/google-announces-gemini-3-5-transcribe-for-ai-powered-speech-to-text/","name":"Ars Technica AI","title":"Google announces Gemini 3.5 Transcribe for AI-powered speech-to-text","records":1},{"url":"https://campustechnology.com/articles/2026/08/26/meta-releases-muse-glimmer-an-open-weight-ai-model-designed-to-run-on-consumer-hardware.aspx?admgarea\\\\u003dnews","name":"Google News AI Model Launch Tracker","title":"Meta Releases Muse Glimmer, an Open-Weight AI Model Designed to Run on Consumer Hardware -- Campus Technology","records":1},{"url":"https://www.marktechpost.com/2026/08/26/z-ai-releases-glm-5-3-flash-a-320b-a18b-natively-multimodal-moe-with-a-1m-token-context/","name":"MarkTechPost","title":"Z.ai Releases GLM-5.3-Flash: A 320B-A18B Natively Multimodal MoE With a 1M-Token Context","records":1},{"url":"https://campustechnology.com/articles/2026/08/26/meta-releases-muse-glimmer-an-open-weight-ai-model-designed-to-run-on-consumer-hardware.aspx","name":"Google News AI Model Launch Tracker","title":"Meta Releases Muse Glimmer, an Open-Weight AI Model Designed to Run on Consumer Hardware -- Campus Technology","records":1},{"url":"https://siliconangle.com/2026/08/26/z-ai-open-sources-ox-alpha-model-as-glm-5-3-flash/","name":"Google News AI","title":"Z.ai open-sources ‘Ox Alpha’ model as GLM-5.3-Flash - SiliconANGLE","records":1},{"url":"https://www.cnbc.com/2026/08/27/zai-shares-surge-new-ai-model-using-chinese-chips.html","name":"Google News AI","title":"Z.ai shares surge 8% after releasing new AI model running only on Chinese chips","records":1},{"url":"https://invezz.com/news/2026/08/27/alibaba-unveils-cheaper-qwen-ai-model-is-this-what-its-stock-needed/","name":"Google News AI Model Launch Tracker","title":"Alibaba unveils cheaper Qwen AI model: is this what its stock needed?","records":1},{"url":"https://www.deccanchronicle.com/technology/alibaba-releases-smaller-cost-effective-qwen-ai-model-1982689","name":"Google News AI Model Launch Tracker","title":"Alibaba Releases Smaller, Cost-Effective Qwen AI Model","records":1},{"url":"https://the-decoder.com/googles-gemini-3-5-transcribe-turns-speech-to-text-in-85-languages-while-auto-correcting-your-verbal-stumbles/","name":"The Decoder","title":"Google's Gemini 3.5 Transcribe turns speech to text in 85 languages while auto-correcting your verbal stumbles","records":1},{"url":"https://the-decoder.com/the-chinese-ai-model-glm-5-3-flash-runs-without-nvidia-and-costs-a-fraction-of-what-the-competition-does/","name":"The Decoder","title":"GLM-5.3-Flash matches top models at a fraction of the cost, and runs without Nvidia","records":1},{"url":"https://itbrief.news/story/leiolai-launches-ai-model-that-runs-on-user-devices","name":"Google News AI Model Launch Tracker","title":"Leiolai launches AI model that runs on user devices","records":1},{"url":"https://x.ai/news/grok-4-6","name":"xAI Official","title":"Introducing Grok 4.6","records":1},{"url":"https://x.ai/news/grok-4-6-amazon-bedrock","name":"xAI Official","title":"Grok 4.6 on Amazon Bedrock","records":1},{"url":"https://openai.com/index/accelerating-life-sciences-research-with-retro-biosciences/","name":"OpenAI News","title":"Accelerating life sciences research","records":1},{"url":"https://openai.com/index/introducing-gpt-live/","name":"OpenAI News","title":"Introducing GPT-Live","records":1},{"url":"https://arstechnica.com/ai/2026/06/anthropic-says-these-topics-are-too-dangerous-to-let-its-fable-5-model-talk-about/","name":"Ars Technica AI","title":"Anthropic says these topics are too dangerous to let its Fable 5 model talk about","records":1},{"url":"https://techcrunch.com/2026/06/09/anthropics-fable-5-can-make-weirdly-fun-video-games-with-the-click-of-a-button/","name":"TechCrunch AI","title":"techcrunch.com","records":1},{"url":"https://www.artificialintelligencemadesimple.com/p/anthropics-claude-mythos-launch-is","name":"artificialintelligencemadesimple.com","title":"Anthropic's Claude Mythos Launch Is Built on Misinformation","records":1},{"url":"https://www.anthropic.com/glasswing","name":"anthropic.com","title":"Project Glasswing: Securing critical software for the AI era","records":1},{"url":"https://openai.com/index/new-chatgpt-images-is-here/","name":"OpenAI News","title":"The new ChatGPT Images is here","records":1},{"url":"https://www.technologyreview.com/2026/01/26/1131728/inside-openais-big-play-for-science/","name":"MIT Technology Review AI","title":"Inside OpenAI’s big play for science","records":1},{"url":"https://openai.com/index/introducing-gpt-5-2/","name":"OpenAI News","title":"Introducing GPT-5.2","records":1},{"url":"https://www.semafor.com/article/08/16/2026/new-chinese-model-adds-to-ai-competition","name":"Google News AI","title":"New Chinese model adds to AI competition","records":1}],"provenance":{"publisher":"Superpower Daily","source_count":40,"source_urls":["https://openai.com/index/gpt-5-6","https://siliconangle.com/2026/08/24/thomson-reuters-launches-proprietary-ai-model-for-legal-work/","https://thelogic.co/news/thomson-reuters-custom-ai-launch/","https://www.stocktitan.net/news/TRI/thomson-reuters-leverages-its-world-class-data-assets-to-launch-its-6yterby5xq24.html","https://the-decoder.com/thomson-reuters-bets-40m-on-owning-its-ai-instead-of-renting-from-openai-or-anthropic/","https://www.marktechpost.com/2026/08/24/generalist-ai-releases-gen-1-5-a-robot-foundation-model-that-learns-new-tasks-from-one-3-12-second-demo/","https://vir.com.vn/hidreamai-releases-hidream-o1-world-interactive-artificial-intelligence-model-159321.html","https://www.readthepeak.com/p/thomson-reuters-launches-its-own-ai-model","https://the-decoder.com/ibm-drops-open-weight-granite-4-2-family-with-built-in-agentic-capabilities-under-apache-2-0/","https://arstechnica.com/ai/2026/08/ibms-new-granite-4-2-models-ride-the-wave-of-interest-in-local-llms/","https://www.yahoo.com/news/articles/alibabas-qwen-launches-qwen3-8-150149621.html","https://m.economictimes.com/tech/artificial-intelligence/alibabas-qwen-launches-qwen3-8-flash-ai-model-with-lower-training-costs/articleshow/133544818.cms","https://the-decoder.com/alibaba-releases-qwen3-8-flash-next-targeting-ultimate-cost-efficiency/","https://www.marktechpost.com/2026/08/26/alibabas-qwen-team-releases-qwen3-8-flash-next-a-125b-multimodal-moe-with-6b-active-parameters-previewing-the-qwen4-architecture/","https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-3-5-transcribe/","https://cryptobriefing.com/zhipu-glm-53-flash-chinese-ai-chips/","https://autoclaw.z.ai/models/","https://arstechnica.com/ai/2026/08/google-announces-gemini-3-5-transcribe-for-ai-powered-speech-to-text/","https://campustechnology.com/articles/2026/08/26/meta-releases-muse-glimmer-an-open-weight-ai-model-designed-to-run-on-consumer-hardware.aspx?admgarea\\\\u003dnews","https://www.marktechpost.com/2026/08/26/z-ai-releases-glm-5-3-flash-a-320b-a18b-natively-multimodal-moe-with-a-1m-token-context/","https://campustechnology.com/articles/2026/08/26/meta-releases-muse-glimmer-an-open-weight-ai-model-designed-to-run-on-consumer-hardware.aspx","https://siliconangle.com/2026/08/26/z-ai-open-sources-ox-alpha-model-as-glm-5-3-flash/","https://www.cnbc.com/2026/08/27/zai-shares-surge-new-ai-model-using-chinese-chips.html","https://invezz.com/news/2026/08/27/alibaba-unveils-cheaper-qwen-ai-model-is-this-what-its-stock-needed/","https://www.deccanchronicle.com/technology/alibaba-releases-smaller-cost-effective-qwen-ai-model-1982689","https://the-decoder.com/googles-gemini-3-5-transcribe-turns-speech-to-text-in-85-languages-while-auto-correcting-your-verbal-stumbles/","https://the-decoder.com/the-chinese-ai-model-glm-5-3-flash-runs-without-nvidia-and-costs-a-fraction-of-what-the-competition-does/","https://itbrief.news/story/leiolai-launches-ai-model-that-runs-on-user-devices","https://x.ai/news/grok-4-6","https://x.ai/news/grok-4-6-amazon-bedrock","https://openai.com/index/accelerating-life-sciences-research-with-retro-biosciences/","https://openai.com/index/introducing-gpt-live/","https://arstechnica.com/ai/2026/06/anthropic-says-these-topics-are-too-dangerous-to-let-its-fable-5-model-talk-about/","https://techcrunch.com/2026/06/09/anthropics-fable-5-can-make-weirdly-fun-video-games-with-the-click-of-a-button/","https://www.artificialintelligencemadesimple.com/p/anthropics-claude-mythos-launch-is","https://www.anthropic.com/glasswing","https://openai.com/index/new-chatgpt-images-is-here/","https://www.technologyreview.com/2026/01/26/1131728/inside-openais-big-play-for-science/","https://openai.com/index/introducing-gpt-5-2/","https://www.semafor.com/article/08/16/2026/new-chinese-model-adds-to-ai-competition"],"methodology":["Each verified model-launch record contributes one release observation.","Developer names are normalized during extraction; series preserve currently disclosed attribution.","Release volume measures activity, not capability or adoption."],"snapshot_hash":"e7fb6b6162760400557eb0124b9c63722247fe8c38df5f7676e189360138078a"},"distributions":{"csv":"https://superpowerdaily.com/api/research/release-velocity/versions/v7?format=csv","json":"https://superpowerdaily.com/api/research/release-velocity/versions/v7?format=json"}}