--- title: DeepSeek and Moonshot trained new models on GPT-6 Astra,... description: DeepSeek and Moonshot AI relied heavily on OpenAI's GPT-6 Astra to train their latest models, The Information reported, citing people with direct knowledge of... canonical: https://daily.dev/posts/deepseek-and-moonshot-trained-new-models-on-gpt-6-astra-report-says-cjsotuouy twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: DeepSeek and Moonshot trained new models on GPT-6 Astra, report says | daily.dev og:description: DeepSeek and Moonshot AI relied heavily on OpenAI's GPT-6 Astra to train their latest models, The Information reported, citing people with direct knowledge of... og:url: https://daily.dev/posts/deepseek-and-moonshot-trained-new-models-on-gpt-6-astra-report-says-cjsotuouy og:image: https://api.daily.dev/og/posts/CJSotUouY.png og:image:alt: DeepSeek and Moonshot trained new models on GPT-6 Astra, report says og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # DeepSeek and Moonshot trained new models on GPT-6 Astra, report says **[The Next Web](https://daily.dev/sources/tnw)** · 2 min read · 1 upvotes · 0 comments ## Summary DeepSeek and Moonshot AI relied heavily on OpenAI's GPT-6 Astra to train their latest models, The Information reported, citing people with direct knowledge of the matter. The models named are Moonshot's Kimi K3.1 Pro, released last week, and an updated version of DeepSeek's V4.1-Flash, launched on 17 October. Distillation lets a developer send prompts to a stronger model and train its own system on the responses. OpenAI said it detected related activity involving more than 15,000 users, tied to people linked to the Kimi developer, and shut down the campaign by 28 July. Anthropic said new safeguards cut data collected from Claude in distillation attacks by around 90% to 99%. Moonshot said Kimi K3.1 Pro outperformed GPT-6 Astra on a coding benchmark, though neither company's results are independently verified. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### What is model distillation in AI training? Distillation is a process where developers send large numbers of prompts to a more capable AI model, collect its answers, and use those responses to train their own system. This lets a smaller model reproduce some of a more advanced model's abilities at a fraction of the training cost, without building those capabilities from scratch. _Anyone evaluating how competing AI models get built can track distillation developments on daily.dev._ ### Did OpenAI accuse Moonshot AI of improperly extracting its model's reasoning? Yes, OpenAI said that on 30 September it found people linked to Moonshot's Kimi developer ran a coordinated campaign in July to extract the hidden reasoning behind its models' answers. OpenAI detected activity involving more than 15,000 users, shut down the campaign by 28 July, banned the accounts, and shared its findings with other AI labs. _Developers weighing trust in rival AI labs can follow this dispute's fallout on daily.dev._ ### How much did Anthropic reduce data leakage from Claude distillation attacks? Anthropic said new safeguards reduced the amount of data collected from its Claude models in distillation attacks by around 90% to 99%. This was announced as part of broader industry efforts by labs like OpenAI, Anthropic and Google to make it harder for rivals to replicate their models' capabilities through distillation. _Teams relying on Claude's outputs can keep up with these safeguard changes on daily.dev._ ## Similar posts on daily.dev - [DeepSeek V4, Kimi K3, and the Strange New Price of Intelligence](https://daily.dev/posts/deepseek-v4-kimi-k3-and-the-strange-new-price-of-intelligence-ugfsk7yvx) · Medium · 1 upvotes · 0 comments --- Tags: [#machine-learning](https://daily.dev/tags/machine-learning), [#llm](https://daily.dev/tags/llm), [#openai](https://daily.dev/tags/openai), [#deepseek](https://daily.dev/tags/deepseek), [#moonshot-ai](https://daily.dev/tags/moonshot-ai) [View this post on daily.dev](https://daily.dev/posts/deepseek-and-moonshot-trained-new-models-on-gpt-6-astra-report-says-cjsotuouy) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"DeepSeek and Moonshot trained new models on GPT-6 Astra, report says","url":"https://daily.dev/posts/deepseek-and-moonshot-trained-new-models-on-gpt-6-astra-report-says-cjsotuouy","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/deepseek-and-moonshot-trained-new-models-on-gpt-6-astra-report-says-cjsotuouy"},"datePublished":"2026-10-29T09:07:36.828Z","dateModified":"2026-10-29T09:08:04.184Z","description":"DeepSeek and Moonshot AI relied heavily on OpenAI's GPT-6 Astra to train their latest models, The Information reported, citing people with direct knowledge of...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/f1b9675d1cd119d714d92ba300526cd4?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/f1b9675d1cd119d714d92ba300526cd4?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"The Next Web","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"The Next Web","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/tnw","url":"https://daily.dev/sources/tnw"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/deepseek-and-moonshot-trained-new-models-on-gpt-6-astra-report-says-cjsotuouy","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":1},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"machine-learning,llm,openai,deepseek,moonshot-ai","timeRequired":"PT2M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"The Next Web","item":"https://daily.dev/sources/tnw"},{"@type":"ListItem","position":3,"name":"DeepSeek and Moonshot trained new models on GPT-6 Astra, report says"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/deepseek-and-moonshot-trained-new-models-on-gpt-6-astra-report-says-cjsotuouy#faq","mainEntity":[{"@type":"Question","name":"What is model distillation in AI training?","acceptedAnswer":{"@type":"Answer","text":"Distillation is a process where developers send large numbers of prompts to a more capable AI model, collect its answers, and use those responses to train their own system. This lets a smaller model reproduce some of a more advanced model's abilities at a fraction of the training cost, without building those capabilities from scratch. Anyone evaluating how competing AI models get built can track distillation developments on daily.dev."}},{"@type":"Question","name":"Did OpenAI accuse Moonshot AI of improperly extracting its model's reasoning?","acceptedAnswer":{"@type":"Answer","text":"Yes, OpenAI said that on 30 September it found people linked to Moonshot's Kimi developer ran a coordinated campaign in July to extract the hidden reasoning behind its models' answers. OpenAI detected activity involving more than 15,000 users, shut down the campaign by 28 July, banned the accounts, and shared its findings with other AI labs. Developers weighing trust in rival AI labs can follow this dispute's fallout on daily.dev."}},{"@type":"Question","name":"How much did Anthropic reduce data leakage from Claude distillation attacks?","acceptedAnswer":{"@type":"Answer","text":"Anthropic said new safeguards reduced the amount of data collected from its Claude models in distillation attacks by around 90% to 99%. This was announced as part of broader industry efforts by labs like OpenAI, Anthropic and Google to make it harder for rivals to replicate their models' capabilities through distillation. Teams relying on Claude's outputs can keep up with these safeguard changes on daily.dev."}}]} ```