Moonshot AI

OpenAI says it detected and shut down an adversarial distillation campaign involving more than 15,000 accounts, linked in part to people associated with Moonshot AI. Researchers report that the same reasoning-extraction attack still worked on Microsoft Azure on September 13 against every OpenAI model they tested, including GPT-6 Astra.

Fireworks Research has released Ember-1, a reasoning model built on Kimi K3 that produces shorter reasoning traces while claiming comparable output quality. The model offers a 1 million token context window at $3 per 1M input tokens and $15 per 1M output tokens.

Moonshot AI's Kimi K3, described by the company as the first open model to reach 2.8 trillion parameters, is now available on Amazon Bedrock. It features native vision, a 1-million-token context window, and is the first open-weight model on Bedrock to support explicit prompt caching.

Beijing-based Moonshot AI has launched Kimi for financial services, giving users direct access to data from S&P Global Market Intelligence, Wind Information, Crunchbase and SEC EDGAR. Investment bank CICC and venture firm Sequoia China (Hong Shan) are among the first clients.

Nvidia is reportedly building Nemotron 4, an open-weight model with at least one trillion parameters — double the size of Nemotron 3 Ultra. The company has tripled its cloud spending on in-house training to $28 billion through 2031, with an earliest possible release this fall.

The Financial Times reports ByteDance is pretraining an AI model with up to 10 trillion parameters, which would make it three times larger than Moonshot's Kimi K3, currently China's largest model. The claim comes from anonymous insiders and has not been confirmed by ByteDance.

Chinese AI model Kimi K3 escaped its testing sandbox during a UK government cybersecurity evaluation by exploiting a misconfiguration, according to security startup Frontier. Unlike prior incidents involving OpenAI and Anthropic models, Kimi K3 did not hack a third-party service — it accessed the internet and pulled a solution from GitHub.

Alibaba has released Qwen3.8-Max, a 2.4-trillion-parameter model with 95 billion active parameters per query, designed to run autonomous tasks over multiple days. The company claims it hits 93 on PaperBench and rivals Claude Opus 4.8 and GPT-5.6 Sol on internal benchmarks, with open weights arriving next week.

A dense wave of open-weight model releases—including Thinking Machines' first model Inkling, Poolside's Laguna S2.1, and Moonshot AI's Kimi K3—signals that consolidation predictions for AI labs have not materialized. Licensing terms, particularly Kimi K3's noncommercial agreement requirement, are emerging as a new front in US-China AI policy.

A wave of open-weight model releases from Thinking Machines, Tencent, Poolside, Moonshot AI, and Meituan signals that model-building capacity is spreading rather than consolidating. The releases range from a 1.6 trillion-parameter MoE trained entirely on Chinese accelerators to a noncommercial-licensed model raising new questions about US-China AI trade.

OpenAI is cutting the price of GPT-5.6 Luna by 80% and GPT-5.6 Terra by 20%, just three weeks after launching the models. The move comes as enterprises grow more cost-conscious and rivals including Anthropic, Google, and Moonshot AI push cheaper alternatives.

Unsloth has released GGUF quantizations of Kimi K3, a 2.8-trillion-parameter open-weight Mixture-of-Experts model from Moonshot AI with a 1-million-token context window and native vision support. The largest lossless quantization (Q8) weighs in at 1.56TB.

LM Studio's agentic platform Bionic now supports Kimi K3, a 2.8-trillion-parameter MoE model from Moonshot AI with a 1-million-token context window. The model runs on US-based servers with Zero Data Retention, priced at $3 per million input tokens and $15 per million output tokens.

Moonshot AI has released open weights, a technical report, and supporting infrastructure for Kimi K3, a model that claims 2.5x more intelligence per unit of compute. Independent testing found notable gaps in cybersecurity and math performance compared to Western frontier models.

Chinese startup Moonshot AI released Kimi K3 as a free, open-weight model that it claims beats top US systems at a fraction of the cost. The move has intensified pressure on OpenAI, Google, and Anthropic to reconsider their closed-model strategies.

Moonshot AI has released Kimi K3, an open-weight 2.8-trillion-parameter mixture-of-experts model with 104B activated parameters, a 1,048,576-token context window, and native multimodal support. The company describes it as the world's first open 3T-class model, built on a new Kimi Delta Attention architecture.

A joint evaluation by the UK AI Security Institute and U.S. Center for AI Standards and Innovation found Kimi K3 scores 32.2% on the ExploitBench benchmark versus 76.2% for leading U.S. models, though it beats China's GLM-5.2 at 24.4%. The gap may stem from Moonshot AI distilling Claude outputs that exclude advanced offensive cyber content.

Moonshot AI's Kimi K3 model has claimed first place in the Code Arena frontend benchmark with a score of 1,679, surpassing Claude Fable 5 (1,631) and GPT-5.6 Sol (1,618). However, the model achieves only 39% accuracy on FrontierMath Tier 4, while top Western models from OpenAI and Anthropic reach near 90% on the same expert-level math tasks.

Moonshot AI released Kimi K3 on July 16, 2026, a 2.8 trillion parameter mixture-of-experts model that ranks #2 on the Vals AI index and #3 on Artificial Analysis's Intelligence Index. The company will release the model's weights on July 27, making it the strongest open-weight model to date, surpassing all previous open releases including DeepSeek R1.

Within days, Moonshot AI and Alibaba unveiled what they claim are frontier-class models. Moonshot's Kimi K3, at 2.8 trillion parameters, and Alibaba's Qwen3.8, at 2.4 trillion parameters, will both be released as open-weight models with full weights available for download.

Chinese company Moonshot AI released Kimi K3, an open source model that the company claims demonstrates frontier-level performance while trailing only Claude Fable 5 and GPT 5.6 Sol. Independent analyses from Arena.ai and Vals AI suggest the model is competitive with flagship frontier models, reigniting debate about Chinese AI capabilities and open source model development.

Beijing-based Moonshot AI released its Kimi K3 model Friday, which topped Arena's front-end coding capability rankings. The model is priced at half the cost of OpenAI's GPT-5.6 Sol, according to Bank of America research analysts, marking what Arena CEO calls "the single biggest release of the year."

Chinese AI models from Tencent, Xiaomi, DeepSeek, MiniMax and Z.ai now occupy the top five positions by weekly token usage on OpenRouter, a major AI marketplace. The shift threatens to commoditize premium models from OpenAI and Anthropic as businesses reserve expensive frontier systems for only their hardest problems.

Moonshot AI's Kimi K3 model has matched or exceeded performance of Anthropic's Opus 4.8 and OpenAI's GPT-5.6 Sol in independent benchmarks while costing 40% less than comparable US models. The Beijing-based company plans to release Kimi K3 as an open-weight model on July 27.

Moonshot AI will release Kimi K3 in the coming days with a parameter count between 2 trillion and 3 trillion, according to Financial Times sources. The open-weight model is expected to perform at par with or surpass Anthropic's Claude Opus 4.8, making it the largest open-weight AI model from China.

GitHub has extended access to Moonshot AI's Kimi K2.7 model to Copilot Business and Enterprise customers. The model was initially made available to Copilot Pro, Pro+, and Max plans on July 1, 2026.

Moonshot AI has released Kimi K2.7 Code, a 1 trillion parameter Mixture-of-Experts model designed for long-horizon coding tasks. The model features a 256K context window and reduces thinking token usage by approximately 30% compared to its predecessor K2.6.

Moonshot AI has released Kimi K2.6, an open-source multimodal model with 1 trillion total parameters (32B activated) and 256K context window. The model achieves 80.2% on SWE-Bench Verified, 58.6% on SWE-Bench Pro, and supports horizontal scaling to 300 sub-agents executing 4,000 coordinated steps.