Here are today's news for 1st of November about Anthropic, Moonshot AI, Mistral AI and OpenAI. Let's dive into the details. On October 26, Anthropic publicly filed an S, 1 registration statement for a proposed IPO, planning to list Class A shares on Nasdaq under “ANT”; it has not set the share count or price range. The filing follows a confidential submission in June and names Goldman Sachs and Morgan Stanley as lead underwriters. It reveals striking growth alongside steep costs: revenue reached $16.29 billion in the first half of 2026, versus $1.16 billion a year earlier, while operating loss narrowed to $693 million. Net loss was $80.7 billion, largely reflecting a $72.4 billion noncash mark, to, market charge on financing instruments. Anthropic also disclosed $417 billion in future compute commitments, underscoring the infrastructure burden behind its Claude business. More than 1,900 customers generated at least $1 million each over the 12 months to September 30. Moonshot AI introduced Kimi K3.1 Pro on October 26, positioning it as the first model in a new series and its most capable open, weight model in the 3, trillion, parameter class. The model retains Kimi K3’s reported mixture, of, experts design, 2.8 trillion total parameters, about 104 billion active, and supports a context window of up to 1,048,576 tokens. Moonshot says post, training improvements target coding, financial research, 3D creation and long, running agent tasks; on its internal Kimi Code Bench 3.0 long, horizon test, the score rose to 61.2 from K3’s 55.95. API pricing fell to $2 per million uncached input tokens and $10 per million output tokens, from K3’s $3 and $15. Artificial Analysis gave K3.1 Pro an Intelligence Index score of 55, fourth in its comparison, and estimated $2.11 per index task. The upgrade builds on K3, launched in July with a million, token context, and intensifies competition with US frontier models. Moonshot says it will publish K3. On October 29, Mistral AI released the open weights for Mistral Large 4, nicknamed “Le Chonk,” a natively multimodal model with about 1 trillion parameters and 52 billion active per task. The French company says the upgraded model, trained for three additional weeks after its October 6 preview, reached 41% on Terminal, Bench 4.0, up from 28%, and uses 47% fewer tokens per task than China’s GLM, 5.3 on the Artificial Analysis Coding Agent Index, where both score 48%. It scored 51% on the Artificial Analysis Cyber Index. Weights are downloadable from Hugging Face under a modified, MIT, style license; commercial model, as, a, service operators need a separate license. Mistral’s API price is $0.70 per million input tokens and $2.10 per million output tokens. Self, hosting gives businesses and governments more control over data and deployment, while strengthening Europe’s open, weight alternative to US and Chinese labs. Mistral says the model was trained in its European datacentres on 3,800 NVIDIA Grace Blackwell GPUs. OpenAI released GPT, 6 Rosalind Research on October 30, 2026, bringing GPT, 6 reasoning to multistep biology, drug, discovery and translational, medicine work. It is rolling out to approved research organizations through a trusted, access program, in Work, Codex and the API, not ChatGPT’s Chat tab; the API model name is `gpt, 6, rosalind`. OpenAI says Rosalind uses the same weights as GPT, 6 Astra but has fewer refusals on dual, use biology queries, while retaining safeguards against assistance that could enable biological weaponization. The company rates its biological and chemical capability High, below Critical, and retains Astra’s Critical cybersecurity rating. On OpenAI’s internal tests, it scored 49.6% on MedChemBench v7 (99 tasks), versus 28.9% for GPT, 5.5 Rosalind; 37.7% on GeneBench Pro v13 (129 tasks), versus 16.7%; and 61.3% on LifeSciBench Gold (94 tasks), versus 44.0%. These small, company, run evaluations are not proof of drug success. The launch extends OpenAI’s existing Rosalind life, sciences offering: GPT, 5. On Nov. 1, 2026, Microsoft made its National AI Safety & Privacy Standard available for U.S. school districts through new or existing contracts, without requiring renewal or renegotiation. Negotiated with the American Federation of Teachers and United Federation of Teachers, the 10, part standard bars using student and educator data to train or improve AI models, sell it, or use it for advertising. A narrow exception allows processing for safety and security, including detecting self, harm risks and threats of violence. The terms also address tracking, human oversight, transparency, cybersecurity and data retention; breach reporting is required within 72 hours. The commitments become contractually enforceable when adopted, giving districts stronger leverage over Microsoft education AI than ordinary policy promises. They affect schools using covered generative AI services, but do not prohibit classroom AI. That's all for today's AI news briefing. Have a great day!