Search “Search”

Moonshot AI releases Kimi K3.1 Pro with 3-trillion-level parameters and a 1-million-token context

Moonshot AI has released Kimi K3.1 Pro, a new-generation open-source large model with 3-trillion-level parameters that supports a 1-million-token context window, retaining its title as the open-source model with the largest parameter scale in the world. The model is a post-trained version of Kimi K3, with a substantial increase in post-training compute, and it shows notable improvements in coding, 3D creation, financial research and long-horizon Agent capabilities; across multiple international benchmarks, its overall intelligence level is now approaching that of the world's leading closed-source models. The report also noted that OpenRouter statistics show global total AI large-model API calls from October 12 to 18 reached 164 trillion tokens, and that the weekly call volume of Chinese large models has exceeded that of the United States for 25 consecutive weeks.

Moonshot AI Releases Kimi K3.1 Pro: A 3T-Class Open-Weight Model Strengthening Coding, 3D and Finance Agents

Moonshot AI has released Kimi K3.1 Pro, calling it the strongest 3T-class open-weight model, featuring a 1M-token context and multimodal capabilities. It simultaneously launched a Highspeed version that is about 2–3× faster, with full weights planned to be open-sourced before 2026-11-04. The model scores 65.7% on Terminal-Bench 4.0, above GPT-6 Astra's 58.2% and slightly below Claude Opus 5.5's 66.4%; it scores 92.7% on BenchCAD with tools and 64.4% on OfficeQA Pro v2. The open-weight version removes cyber-related data from both mid-training and RL, and its xdev-bench-hard score drops from 90.5% to 81.0%.

Moonshot AI releases Kimi K3.1 Pro: a 3T-class open-weight model with a 1 million token context

Moonshot AI releases Kimi K3.1 Pro, the first model in the K3.1 series and its strongest 3T-class open-weight model to date, with a 1 million token context window, alongside Kimi K3.1 Pro Highspeed, which is 2-3 times faster. The model is now available on Kimi.com, Kimi Work, Kimi Code and the Kimi API, with full model weights to be released on November 4. On Terminal-Bench 4.0, K3.1 Pro scores 65.7%, higher than GPT-6 Astra's 58.2% and slightly below Claude Opus 5.5's 66.4%; on the internal benchmark Kimi Code Bench 3.0 Long-horizon it scores 61.2%, higher than K3's 55.95%.

Google releases Gemini 4 Argon, showcasing 5 community creative projects

Google releases new frontier model Gemini 4 Argon, stating it performs outstandingly on complex tasks such as coding, mathematics, and 3D world building, and surpasses other frontier models on benchmarks across multiple domains. The official showcase features 5 community projects, including a 3D Earth that maps wildfires and seismic activity in real time, a turbofan engine simulation with real-time data, a 3D carriage generated from a single image, and a multiplayer capture-the-flag browser game. The model has been widely available since last week.

Hugging Face Releases Open-Source Model Carbon-X, Which Can Analyze and Generate DNA Sequences Using Natural Language

Hugging Face has released the open-source model Carbon-X, combining DNA analysis and generation with natural language so that researchers can ask questions directly and receive predictions and explanations. The model is based on Qwen3-14B and was trained on genomic sequences and natural language from humans, other mammals, plants, and fungi, with viruses explicitly excluded from the training data. Researchers can run it locally, avoiding the sharing of sensitive patient data with external AI companies, and can also generate new DNA sequences to assist gene therapy research.

Inception Releases Diffusion Language Model Mercury 3, Claiming Speeds Four to Eight Times Faster Than GPT-6 Luna

Inception has released the diffusion language model Mercury 3, claiming it wins 8 of 12 benchmarks and runs four to eight times faster than GPT-6 Luna and Claude Haiku 5.5. In Inception's own testing, Mercury 3 scored 76.84 on IFBench, 15 points higher than GPT-6 Luna; it scored 51.08 on SWE-Bench Atlas, versus 33.33 and 32.26 for its competitors. Claude Haiku 5.5 led on the DeepSWE long-code test and long-text retrieval. The model generates 1,068 tokens per second on Nvidia GPUs, is priced at $0.20 per million input tokens and $0.70 per million output tokens, with a 50% discount during the launch period, and is now available on the Inception platform as well as OpenRouter, Baseten, and Vercel; there are currently no independent tests.

NVIDIA Releases Nemotron 3.5 Series of Open Models, Covering Embedding, Reranking, and Document Parsing

On October 21, 2026, NVIDIA expanded the Nemotron Retriever series, launching Nemotron 3.5 Embed (8B BF16, 1B BF16, 1B NVFP4), Nemotron 3.5 Rerank 8B BF16, and the 631-million-parameter Nemotron 3.5 Parse, aimed at multimodal retrieval and document processing for enterprise agent workflows. NVIDIA says that adding Rerank to the agent retrieval pipeline reduces the token cost per query by about 14% (with the 8B embedding model) and 13% (with the 1B model), respectively, while also improving retrieval accuracy. Parse scores 75.20 on ParseBench and processes 10.5 pages per second on a single H100; the open weights, data, and training recipes have been released on Hugging Face and can be deployed via NVIDIA NIM as well as Baseten, DeepInfra, and FriendliAI.

MiniMax Releases M3.1 Flash, Scoring 29.6% on Agents' Last Exam to Rank First Among Chinese Models

MiniMax has officially released M3.1 Flash, the model previously known as Space Bunny (Yutu) that topped the charts anonymously on OpenRouter. It scores 29.6% on Agents' Last Exam, ranking first among Chinese models. The model achieves 73.7% on DeepSWE v1.1, 82.2 points on CyberGym, and 61.7 points on OSWorld V2, with output speeds of up to 200+ TPS and a P50 maintained at 170+ TPS. Pricing is 1 yuan per million input tokens, 4 yuan for output, and 0.1 yuan for cache reads.

Memory Tensor Open-Sources MemT 1.0 Series of Memory-Native Models

Memory Tensor has open-sourced the MemT 1.0 series of memory-native models, comprising two variants: Nano with 36B parameters and 3B activated, and Flash with 288B parameters and 17B activated, built on the Qwen3.6-35B-A3B and DeepSeek-V4-Flash bases, respectively. The models add a memory branch parallel to attention within selected Transformer modules, encoding experience into an updatable matrix state that participates in subsequent generation. Across 500 SWE-bench Verified tasks, Nano achieved an average success rate of 69.73% at key checkpoints, versus a no-experience baseline of 66.07%; under a 128K long history, end-to-end inference speed is approximately 1.86 times that of the full-context approach, and the persistent state footprint per trajectory is one sixty-sixth that of the full-context KV Cache.

Gemini 4 Argon Sets New SoTA on the WANDR Benchmark and Launches on Perplexity

Perplexity CEO Aravind Srinivas said that Gemini 4 Argon is a frontier model that has set a new SoTA on the Wide-And-Deep-Research (WANDR) benchmark, and it is now available to all Perplexity users, covering fast search and the Computer agent mode. Perplexity's official account previously said that the model set a new record on WANDR and was offered to Pro and Max subscribers.

NVIDIA Expands Nemotron Model Family with New Super VL and Two Bolt Models

NVIDIA has added three models to the Nemotron open-source model family, covering enterprise workflows, local AI agents, and conversational applications across data centers, desktops, and edge devices. These include the 120-billion-parameter Nemotron 3.5 Super VL, which NVIDIA claims delivers up to 5x the throughput of the next-best comparable model, with accuracy approaching that of its largest reasoning model at roughly one-fifth the parameter count; there are also two small models in the Nemotron 3.5 Bolt series: 74B-A9B 24GB for everyday agent tasks on RTX and DGX Spark, and E4B 4GB for conversational and physical AI agents. Super VL and Bolt 74B-A9B are already available, and Bolt E4B is scheduled for release on October 23.

English translation is being prepared.

English translation is being prepared.

StepFun open-sources Step 5 Preview, a 600B-parameter MoE model that ranks second on the OpenRouter weekly chart

StepFun officially releases the Step 5 Preview model weights, with BF16 and FP8 versions launched simultaneously on Hugging Face and ModelScope. The model uses a sparse MoE architecture, with 600B total parameters and 27B activated parameters, supports a 1 million Token context window, and scores 44 on Artificial Analysis's Intelligence Index, placing it in the global top three open-source models. As of October 15, 2026, its Token usage ranks second on both the OpenRouter weekly and daily charts, with about 18.8 trillion Tokens used in the past seven days; according to evaluation data at release, its average cost per task is $0.71, about 1/8 that of Claude Opus 5.

English translation is being prepared.

English translation is being prepared.

English translation is being prepared.

English translation is being prepared.

Fireworks Research Releases Ember-1: A Token-Saving Specialized Model Based on Kimi K3

Fireworks Research has released Ember-1, a specialized model trained on Kimi K3 that reduces token consumption by about 40% while maintaining equivalent quality. The company says that across 7 benchmarks and production traffic from two customers, K3's reasoning process can be shortened by 35–50% without any drop in accuracy; on Bedside Bench, Ember-1 sits on the Pareto frontier for cost and task completion. The model is available on Serverless as a Research Preview, offered alongside the base Kimi K3, with training support on enterprises' own data also opened up.

GPT-6 Family's First Model Astra Released, Author's Hands-On Test Consumed 4 Billion Tokens

OpenAI released Astra, which the author calls the beginning of the GPT-6 family; it performs outstandingly on ARC-AGI-3 and scored highest on Zapier's AutomationBench, with pricing the same as Fable 5.1. During weekend trials, the author consumed 4 billion tokens and found its capability fluctuates noticeably—it is a showcase model rather than one that reliably does work. Astra in Codex can now continue advancing tasks without waiting for an answer, and OpenAI also claims it has achieved its automated research intern goal, with the next goal being an automated AI researcher by March 2028.