Search “Multimodal”

English translation is being prepared.

English translation is being prepared.

Mistral AI Opens the Mistral Large 4 Model Weights: 1T Parameters, 49 Billion Activated Parameters

On October 29 local time, Mistral AI officially opened the model weights of Mistral Large 4 (codename le Chonk). This is a natively multimodal model with 1 trillion total parameters and 49 billion activated parameters, trained from scratch on 3,800 NVIDIA Grace Blackwell GPUs in its own data center in Europe. The company says it scores 48% on the Artificial Analysis Coding Agent Index, roughly on par with GLM-5.3, but consumes 47% fewer Tokens per task on average. The model is released under a modified MIT license: personal, academic, fine-tuning, and conventional commercial use are free, while cloud providers offering MaaS resale need to obtain a separate commercial license; online API pricing is $0.70 per million input Tokens and $2.10 per million output Tokens.

Ant Group's inclusionAI Open-Sources 2B Visual Forensics Model DefakerOne2

Ant Group's GuangJian team has open-sourced DefakerOne2 on GitHub, a 2B-parameter unified visual forensics model built on the Qwen3.5-2B multimodal backbone, covering AI-generated content, DeepFake faces, document tampering, natural image tampering, face and ID anti-counterfeiting, and AI-generated video detection. The model was trained on approximately 50 million samples, employs a three-stage Tide-Style Tuning process from LoRA to full parameters and back to LoRA, and uses entropy-based test-time routing to select between the video and frame pathways. The team claims that it achieves SOTA on 40 of 53 detection benchmarks and on 11 of 14 localization benchmarks, and that the weights and the RRVBench benchmark have been released on ModelScope.

Mistral Large 4 weights go live on Hugging Face; the company says it leads on CyberGym-E2E and…

After a three-week preview, Mistral AI has released the weights of Mistral Large 4 on Hugging Face. This multimodal model was trained on 3,800 Nvidia accelerators and has a parameter count of one trillion, and it is currently still difficult to run on a gaming PC or a single GPU. Mistral says it leads GLM-5.3 and Kimi K3 on the cybersecurity benchmark CyberGym-E2E, and is on par with GLM-5.3 on the agentic coding benchmark Terminal Bench 4.0, but there is no independent evaluation yet; preview pricing is $1.4 per million input tokens and $4.2 per million output tokens.

Diffusion Intelligence releases Clover-35B-A3B-Preview, a 35-billion-parameter diffusion language model

Diffusion Intelligence releases the diffusion language model Clover-35B-A3B-Preview, which uses a MoE architecture with 35 billion total parameters and about 3 billion activated per token, and supports a thinking mode. On the 49 tasks of the web Agent benchmark WindTunnel, its median execution time was 1.52 seconds, lower than the 3.21 seconds of the Jev + Mercury 2.5 combination, with success rates of 98.6% and 95.9% respectively. On code benchmarks, it scored 73.6 on SWE-bench Verified and 51.7 on Terminal-Bench 2.1, and the team said it will subsequently launch a multimodal version and advance on-device adaptation.

Mistral Releases Open Weights for Mistral Large 4 (Le Chonk)

Mistral has released the open weights for Mistral Large 4 (Le Chonk), a 1-trillion-parameter natively multimodal model that continues reinforcement learning on top of the preview version, with improved coding and defensive cybersecurity capabilities. It scores 19.7% on FrontierCode 2.0 and 48% on the Artificial Analysis Coding Agent Index, roughly on par with GLM-5.3, but uses an average of 47% fewer tokens per task; on Terminal-Bench 4.0 it improves by 13 percentage points to 41%. The model supports three reasoning modes, uses a modified MIT-based Mistral Ecosystem Open-Weight license, and is priced via API at $0.70 per million input tokens, $0.07 for cached input, and $2.10 for output.

Moonshot AI releases Kimi K3.1 Pro: 2.8T total parameters, 104B activated parameters, weights scheduled to be opened on November 4

Moonshot AI releases Kimi K3.1 Pro, a multimodal MoE model with 2.8T total parameters, 104B activated parameters, and a 1 million token context window. The Standard and Highspeed versions are now available on the Kimi website, App, Kimi Work, Kimi Code and API. The company says Highspeed offers 2–3x faster output at twice the price of the Standard version. Artificial Analysis ranks it fourth on the Intelligence Index with a score of 55, behind Opus 5.5's 58; on Arena's Code Arena WebDev it ranks 8th with 1692 points, up nine places from K3. Standard pricing is $2 per million tokens for input, $10 for output, and $0.2 for cached input, one-third lower than K3; the weights have not yet been released, and the company promises to open them on November 4.

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%.

Kimi releases K3.1 Pro open-weight model, 2.8T total parameters, 1M token context

Kimi releases K3.1 Pro, its strongest 3T-class open-weight model, using a MoE architecture with 2.8T total parameters and 104B activated parameters, supporting a 1M token context and multimodality. Officially it is claimed to match the strongest closed-source models on coding, agent, and knowledge benchmarks, and to support full-project game development in Unreal, Unity, and Godot as well as parametric 3D modeling in Blender and CAD, with hallucinations in financial scenarios reduced by 65%. K3.1 Pro and K3.1 Pro Highspeed, which is 2–3 times faster, are already live on the Kimi official website, Kimi Work, Kimi Code, and the Kimi API, with open weights planned for release on November 4, 2026.

Kimi Releases 2.8T Flagship K3.1 Pro, Tops SWE Marathon with a Score of 51.7

Kimi has released its multimodal flagship model K3.1 Pro, with 2.8T total parameters and 104B active parameters (MoE), and a standard 1M-token context window. The author says it tops the SWE Marathon leaderboard with a score of 51.7, overtaking Opus 5.5 and GPT-6 Astra, and the company has also promised to open-source the full weights by November 4. K3.1 Pro and K3.1 Pro Highspeed, which is 2–3× faster, are now live on the Kimi web app, Kimi Work, Kimi Code and the Kimi API.

Kimi Releases K3.1 Pro: A 2.8T Total-Parameter, 104B Active MoE Open-Source Model

Kimi officially released Kimi K3.1 Pro, a 3T-class open-weight model built on a MoE architecture with 2.8T total parameters and 104B active parameters, supporting a 1M-token context and multimodality. The company says it performs comparably to leading models such as Opus 5.5 and GPT-6 Astra on benchmarks while costing significantly less, and that hallucinations in financial modeling scenarios dropped by 65%. Kimi K3.1 Pro and K3.1 Pro Highspeed, which is 2–3× faster, are now available on the Kimi platform, Kimi Work, Kimi Code, and the Kimi API, with open weights to be released on November 4.

Kimi K3.1 Pro launches on OpenRouter, 1M context, 2.8T parameter MoE open-weight model

Kimi K3.1 Pro is now live on OpenRouter; it is a 1M-context multimodal open-weight model post-trained from Kimi K3, targeting long-horizon coding and agent tasks, priced at $2 per million input tokens and $10 per million output tokens. According to Kimi_Moonshot, the model is a 3T-class open-weight model with 2.8T total parameters and 104B activated parameters (MoE), excelling at full-project coding in Unreal, Unity, and Godot and at parametric 3D modeling in Blender and CAD, with financial modeling hallucinations reduced by 65%, and it is said to match the strongest closed-source models on coding, agent, and knowledge benchmarks. Kimi K3.1 Pro and the 2–3x faster K3.1 Pro Highspeed are now available on the Kimi platform, Kimi Work, Kimi Code, and Kimi API, with weights to be opened on November 4, 2026.

Kimi K3.1 Pro released, rises to 8th on Code Arena WebDev

Moonshot AI released Kimi K3.1 Pro, which ranks 8th on Code Arena WebDev with 1,692 points, matching Claude Opus-5 (Max) while costing 70% less, and rising 9 places from Kimi K3 (Max)'s 17th. The model is priced at $2/$10 per million input/output tokens, with a median cost of $1.23 per task, 29% lower than Kimi K3, and its weights will be opened on November 4. Kimi officially calls it a 3T-class open-weight model, featuring a 1M-token context, a 2.8T total parameter/104B active parameter MoE multimodal architecture, and a 65% reduction in hallucination rate.

Moonshot AI Releases Kimi K3.1 Pro, Scoring 65.7% on Terminal-Bench 4.0

Moonshot AI has released Kimi K3.1 Pro, its first updated version, which scores 65.7% on Terminal-Bench 4.0—0.7 points lower than Claude Opus 5.5 and 7.5 points higher than GPT-6 Astra—while its average cost per task is lower than both. The model is a 3T-class open-weight MoE model with 2.8T total parameters and 104B active parameters, supports a 1M-token context and multimodality; in a tool-augmented hallucination stress test, its false-claim rate dropped from 23.4% in K3 to 8.3%. K3.1 Pro and the 2–3× faster Highspeed version are now available in the Kimi app and API, with full weights scheduled for release on November 4.

Kling 4.0 Launches on the fal Platform

Kling 4.0 is now available on the fal platform. It is a new-generation native multimodal video model that improves the stability of complex motion, reference consistency for characters and scenes, and control precision for cinematic storytelling. It is currently available only through fal UX and fal Agent, with API access coming soon.

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.

English translation is being prepared.

English translation is being prepared.

Perplexity Open-Sources Two Decision Models: pplx-decider-v1.2 and flash-v1.2

Perplexity has open-sourced two decision models: pplx-decider-flash-v1.2 has 4B parameters and costs 1 cent per million tokens, making it the strongest 4B model on the HF Decision Index; pplx-decider-v1.2 has 27B parameters and costs 2 cents per million tokens, ranking 2nd on the overall HF Decision Index leaderboard. Both are now live on the Perplexity Decisions API and offer open weights on Hugging Face. Perplexity Developers says this generation of open-weight multimodal decision models improves on both text and vision scores, delivers inference up to 5x faster, and completes all questions in a request in a single forward pass.