Mistral

OpenAI publishes hundreds of math proofs from unreleased frontier model, Mistral and Reflection AI launch open-weight models to rival China, and more!

Mistral has released a preview of Mistral Large 4, a 1 trillion parameter model with 49 billion active parameters trained on its own cluster of 3,800 NVIDIA Grace Blackwell GPUs. The API preview is live with open weights promised by the end of the month, and its Artificial Analysis score of 38 is a huge jump from Mistral Large 3's 9 — though still roughly six months behind the frontier.

Simon Willison's comment on the Mistral Large 4 discussion at Hacker News, in which he takes a jab about saturated benchmarks and actually runs four frontier models on an absurd SVG prompt.

Mistral AI has launched Mistral Large 4 (Le Chonk) in public preview: a granular Mixture of Experts model with 1.05 trillion total parameters, 49B active per token, a 1.6B-parameter vision encoder, and a 1M-token context window, trained from scratch on 3,800 NVIDIA Grace Blackwell GPUs in Mistral's own European datacenters. The API is live at $1.36 per 1M input and $4.18 per 1M output tokens, with cached input at $0.14; weights and license are promised for end of October 2026, so self-hosting is not yet possible. Its strongest results are in cybersecurity (93% Cybench, 82% CyberGym-E2E), where Mistral says several closed frontier models score near zero because they refuse the task.

Most AI coding demos stop at task managers, weather apps, or simple chatbots. For this project, we take on something more demanding: building an enterprise customer-support platform that can investigate complaints, retrieve relevant policies, recommend resolutions, and keep risky actions behind human approval. This gives us a practical way to test Claude Fable 5.1 as […] The post Building an Enterprise AI Customer Support Platform with Claude Fable 5.1 and Claude Code appeared first on Analytics Vidhya.

GPT-6 Astra Ultrafast, running on NVIDIA Blackwell GPUs, is available now in the OpenAI API and to eligible ChatGPT Work and Codex users. Accelerated by inference optimizations through OpenAI’s models that tap into the capabilities of the NVIDIA Blackwell architecture, Ultrafast offers up to 8x faster token generation than the Astra Standard mode. For developers, […]

The risks of AI aren’t what we think they are, as a recent security incident between China and the United States reveals Amid a barrage of news stories warning about superintelligent machines rendering humanity extinct, a CNN story describing the opposite scenario – one in which the US military’s reliance on brittle chatbots almost brought the US into war with China – went mostly unnoticed by the public. The biggest international AI news of the past three weeks was Anthropic engineer Jacob Coxon’s resignation. According to him, OpenAI and Anthropic are “racing straight towards self-improving superintelligence and gambling with our lives”. Coxon’s description of a “terminator” scenario, a machine becoming much smarter than humanity and deciding to wipe us out, captured the public, journali…

A new arXiv paper by Ephraim Atta-Duncan tests whether open-weight language models give the same canonical answer to word problems when quantities are rewritten in numerically equivalent forms such as decimals, fractions, percentages, number words, scientific notation, or exactly converted units. Across 3,600 exact-rational problems and 8,600 prompts spanning five transformation families, five open-weight systems show high canonical accuracy (0.969–0.996) after a fixed syntax audit, but orbit correctness and invariance drop to roughly 0.85–0.98. Much of the apparent strict-parser collapse traces to multiplication-form scientific notation falling outside the evaluator's number grammar, while Mistral Small 4 exhibits a separate semantic failure on unit conversions, with 265 errors off by ex…

Have you ever read a paper in Science or Nature and thought, “Man, that research was so cool. I wish I could try that method on my own data”—only to spend a week wrestling with someone else’s undocumented repo, broken dependencies, and half-finished readme.txt? Well, now you can, more or less. Say hello to Paper2Agent, a new open-source framework that transforms academic reports into interactive AI agents you can talk to. Give it a paper along with the accompanying codebase, data or other supplementary material, and the system automatically extracts the core workflows, then spins up a tested, runnable toolkit that you can use on your own datasets. The concept may sound a little like Google’s NotebookLM (now called Gemini Notebook), which lets you upload documents and chat with an AI about…

arXiv:2609.17538v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly used for structured information extraction from documents, yet their behavior under realistic OCR noise remains poorly understood. We present a systematic benchmark of open-source instruction-tuned LLMs for key-value pair (KVP) extraction under both clean-text and noisy OCR conditions. We evaluate representative decoder-only models (Gemma, Mistral, Qwen2.5, LLaMA 3, and DeepSeek) on the FUNSD, CORD, and SROIE benchmarks using both Gold-text annotations and OCR outputs from PaddleOCR, EasyOCR, and Tesseract. A unified evaluation protocol isolates the effects of input quality, model design, and prompting under consistent conditions. The results show that modern LLMs act as strong semantic extractor…

The French AI lab is using a $3B fundraise to sell control over AI infrastructure, not just model power -- a shift in direction that could matter to U.S. firms in Europe too.

arXiv:2609.09203v1 Announce Type: new Abstract: Existing benchmarks for autonomous AI scientists evaluate only final outputs---generated code, hypotheses, or papers---yet discard the reasoning process by which those outputs were obtained. This makes it impossible to audit scientific methodology, diagnose failure modes, or distinguish systematic reasoning from fortunate guessing. We present \textbf{OpenDiscoveryTrace}, a public dataset of 558 complete AI scientific agent trajectories that captures how models reason, not just what they produce. Each trajectory records a structured 9-field-per-step trace---including thoughts, tool calls, observations, errors, revision triggers, and self-reported confidence---as models execute 124 scientific tasks spanning drug discovery, materials science, g…

This week, Mistral announced it raised €3 billion in a Series D funding round, pushing its post-money valuation past €21 The post Mistral wants open-weight AI to compete at the frontier. It just raised $3.5 billion to do it. appeared first on The New Stack.

Overshadowing Cognition's $48B Series E, Mistral's $24B Series D, Meta's Muse agent, and GPT Image 2.5. The most jam packed, feel the AGI day in the history of AI.

Mistral started as an open-weight AI startup, but its focus is now shifting toward sovereign AI in response to the European market. The new funding round is expected to help bridge that transition.

arXiv:2608.28623v1 Announce Type: new Abstract: Large multimodal reasoning models (LMRMs) are getting increasingly capable, primarily through generating explicit chain-of-thought reasoning before answering. In language models it has been observed that this performance often comes with sycophancy, the tendency of a model to agree with the user over the evidence. However, for LMRMs no reliable method to measure sycophancy yet exists. We bridge this gap by introducing a benchmark and dataset for evaluating LMRM sycophancy when confronted with a wrong answer from a user. Our benchmark pairs four visually grounded datasets spanning mathematical, clinical, temporal, and demographic reasoning with five pressure conditions in single-turn and multi-turn settings. We evaluate sycophancy in the fina…

arXiv:2608.28911v1 Announce Type: new Abstract: The key-value (KV) cache is the dominant memory bottleneck of long-context large language model (LLM) inference, growing linearly with context length. We show that uniform KV quantization on a fractional-bit grid does not degrade gracefully: under a prespecified multi-seed statistical protocol, Llama-3.1-8B-Instruct with an affine quantizer is statistically indistinguishable from FP16 KV down to 2.322 code bits/value and collapses at 2.0 bits - a quality cliff in (2.0, 2.322] that reappears in generation-time quantization and multi-turn dialogue and transfers to Mistral-7B. The cliff reframes importance-aware mixed precision: above it, eight model-internal importance indicators are statistically interchangeable, so the benefit of mixing is g…

See how Gilbert + Tobin combines CEO-led commitment, rigorous governance, and human accountability to scale ChatGPT Enterprise and Codex across the firm.

OpenAI announced ChatGPT Work on July 9th, and have been furiously iterating on it ever since. It is an extraordinarily confusing and very powerful product. Here's what I've figured out about it so far. ChatGPT Work is actually two products The more interesting version of ChatGPT Work is the one that runs in the cloud. This can be accessed via chatgpt.com or through the ChatGPT mobile apps. Let's call it Work Cloud. If you install the ChatGPT desktop app - the app that used to be called Codex - you gain access to a thing called ChatGPT Work that can access files and run programs directly on your computer. Let's call that one Work Local. This one feels more like regular Codex re-skinned to be less intimidating to non-software-developers. For the rest of this article I'm going to talk exclu…

Cohere has released Parse (parse-v5.0), a 2.3B-parameter vision language model that converts PDFs, slides and images into Markdown with HTML tables, bounding boxes and image descriptions. It runs at $1.50 per 1,000 pages through the API, or on dedicated Model Vault instances from $2,500 a month. Cohere reports a ParseBench score of 79.2, ahead of Mistral OCR 4, Azure Document Intelligence and Databricks AI Parse — but that figure averages three of the benchmark's five dimensions and drops charts and visual grounding entirely. The post Cohere Releases Parse 5 (parse-v5.0): A 2.3B Vision Language Model That Turns Enterprise Documents Into Markdown appeared first on MarkTechPost.

The French AI lab extends its push for regional control of AI from Europe to the Middle East.

In this article, you will learn how Gemma 4, Llama 3, and Mistral implement tool calling locally, and what trade-offs each model family presents for...

According to OpenRouter data, agentic AI workloads consume 15x more tokens than a simple chat request. Why? Consider what happens when an AI agent researches a company for an investment decision. The agent queries financial databases, searches news and filings, invokes a sub-agent to run peer comparisons and model valuations, then synthesizes everything into a […]

arXiv:2608.18090v1 Announce Type: new Abstract: Inside a modern language model sits a single internal direction that tracks how positive or negative a sentence feels. We show how to find this valence axis (V-axis) from just 9 emotion category names plus 50 short narrative paragraphs per emotion -- about 1,500 fewer labels than the usual supervised approach -- and that the same direction appears in vision, audio, and human-brain encoders never jointly trained. The recipe: embed nine emotion-anchored story sets in a frozen encoder, take the top principal direction of the nine averaged embeddings. Projecting new inputs onto it captures 93% of supervised performance on SST-2 (Llama-3-8B-Instruct, AUC 0.772 vs. 0.828), correlates with human valence ratings on 11,811 EmoSet images at r=0.636, r…

arXiv:2608.18089v1 Announce Type: new Abstract: Instruction-tuned models often refuse harmful requests in English but comply with the same requests in Yoruba, Igbo, Igala, and Hausa. This suggests that the refusal mechanism is present in the residual stream but fails to activate for low-resource inputs. Recovering it normally requires labelled target-language data and retraining, neither of which is available at scale for most African languages. We introduce Latent Space Refusal Anchoring (LSR-Anchoring), a training-free method that extracts the refusal direction from English prompts and clamps it onto the residual stream at inference time. The primary variant, Mean-Activation Steering (MAS), operates across the four architectures we tested: Llama-3-8B, Llama-3.1-70B, Mistral-7B-Instruct,…

An end-to-end tutorial for supervised fine-tuning of tool-calling LLMs, covering trajectory parsing, structured tool-call extraction, Qwen-compatible ChatML rendering, and LoRA fine-tuning of Qwen3-0.6B on the XYZ-Aquila-SFT dataset.

Mistral AI wants to turn European AI sovereignty from a talking point into a product — one with a service-level agreement attached. The French artificial intelligence company announced Tuesday a three-part expansion of…

arXiv:2608.11210v1 Announce Type: new Abstract: Bayesian calibration of process-based models requires a prior distribution for each model parameter. Despite decades of methodological work, researchers almost always fall back on uniform priors. The main reason is that building informative priors from scientific literature is slow and needs both domain and statistical expertise. We present \textbf{Distribird}, an agentic web application that automates this process. Given a parameter name, physical description, and domain context, Distribird deploys a multi-agent pipeline that searches the literature, extracts and weights reported values by domain relevance, and fits a probability distribution via AIC model selection. When no literature is available, the system falls back to sensible uninfor…

The Paris-based vendor continues to build European AI infrastructure.

That’s a lot of people chatting with their AI friends all day. | Image: Google For the 14th time, a Google product has hit 1 billion users. Google CEO Sundar Pichai posted on X that a billion people are using Gemini every month, and that Gemini is Google's fastest-growing product ever. A billion users is a huge milestone, but Google isn't the first AI app to hit it. OpenAI's ChatGPT hit the mark a few weeks ago, though the company buried the announcement that "more than 1 billion people are putting ChatGPT to work" in an otherwise anodyne blog post about how people use AI. External data suggested ChatGPT crossed 1 billion users as early as this June, but OpenAI hadn't announced anything until that post on August 6th. … Read the full story at The Verge.

Mistral AI has released Shieldstral 1.0 3B, an open-weights, policy-adaptive multimodal safety classifier that frames content moderation as a single yes/no question instead of a fixed harm taxonomy. Operators supply the policy as a plain-language query at inference time and get back a calibrated safety score from one forward pass — no retraining required to re-target the model. Built on Ministral-3-3B-Base-2512 with a Pixtral vision encoder and trained on roughly 54.1M samples, it reports 84.9% average F1 on text safety (matching GPT-OSS-Safeguard-20B), 83.8% on multimodal safety, and 91.3% on Mistral's adaptability benchmark — while fitting in 16GB of VRAM under an Apache 2.0 license. The post Mistral AI Releases Shieldstral 1.0 3B: An Open-Weights Policy-Adaptive Multimodal Safety Class…

arXiv:2608.05162v1 Announce Type: new Abstract: Pooling is a consequential but under-examined design choice in decoder-only concept representation work: practitioners must collapse token-level hidden states into a passage-level vector, yet no shared protocol exists for comparing this choice across concepts, models, and tasks. Reported gains are confounded by simultaneous changes in dataset, layer, construction method, and pooling rule, making principled decisions impossible. We introduce PoolBench, a benchmark that isolates pooling as the experimental variable under a fixed evaluation protocol. PoolBench covers 17 concepts, 19 pooling strategies, and 3 open-weight decoder-only models (Llama-3.1-8B, Gemma-2-9B, Mistral-7B), evaluated on a single audited corpus of 37,693 real-text passages.…

arXiv:2608.04130v1 Announce Type: new Abstract: Vision-language models for autonomous driving primarily rely on cameras and LiDAR, leaving 4D radar largely unexplored as a standalone perceptual modality despite its robustness to adverse visibility and direct measurement of radial velocity. We introduce Radar4D-VLM, a radar-only temporal vision-language model that reasons from ten consecutive 4D-radar point-cloud sweeps without camera or LiDAR input. Radar4D-VLM extracts geometrically grounded object proposals and organizes radar evidence into a compact hierarchy of object, scene, and kinematic tokens. A parameter-efficient projector maps these tokens into frozen language backbones, while auditable prediction heads jointly model object count, spatial distribution, motion state, collision r…

I released LLM 0.32 this morning, the most significant new version of LLM since the initial launch of the project. The new version includes support for visible reasoning traces, server-side provider tools, redesigned content-addressable SQLite logs, new models, and new features enabled by the OpenAI Responses API. I also released new versions of the llm-anthropic, llm-gemini, and llm-openrouter plugins, each with substantial updates of their own. Headline features for LLM CLI users Running LLM against reasoning models now displays their reasoning traces to standard error, so you can see what they are "thinking" without that information being included in the standard output that you might pipe to another tool. Add -R/--hide-reasoning to turn this off. LLM includes support out-of-the-box fo…

The AI Model Hub is a central platform for comparing open-source large language models from leading developers like Meta, Alibaba, Google, and Mistral. It provides detailed specifications for over 100 active models, including context windows, architectures, parameter counts, licenses, and benchmarks.

This paper presents a two-part contribution for large-scale chatbot validation: a methodology for creating high-fidelity synthetic customer agents (SCAs) as digital twins, and an SCA-based validation framework combining automated LLM-as-a-Judge evaluation, human expert testing, and adversarial probing. The approach was validated at a leading UK bank, providing a scalable pathway toward regulatory compliance.

An AI employee is an agent with a persistent workspace and the tools to complete assigned work end to end, far beyond a simple chat assistant. Construct is a work OS that provides files, memory, schedules, and workflows for such agents.

A new study benchmarks the performance cost of enabling confidential computing for LLM inference on an NVIDIA H100 GPU under Intel TDX. Using Mistral-7B and Qwen3-30B-A3B models, results show a 21.8%-27.8% increase in time-to-first-token and 17.7%-21.1% drop in global token throughput in confidential mode. The larger model reaches saturation earlier, highlighting the need for capacity planning adjustments.

AI Maestro orchestrates AI coding agents to work on a task board, turning software delivery into a coordinated multi-agent pipeline rather than a single chat session.

The alliance strengthens Mistral’s position as the leading European AI vendor, while extending Microsoft’s presence in Europe.

NVIDIA Vera Rubin NVL72 production is ramping up with partners CoreWeave, Google Cloud, Microsoft Azure and Oracle Cloud Infrastructure. The platform delivers highest performance per watt and lowest token cost, with 10x more throughput per megawatt than Grace Blackwell NVL72 in benchmarks. It also powers Europe's open-model era through a partnership between Microsoft and Mistral.

A single 24GB GPU is the practical floor for serious local inference. This guide compares six open-weight models that fit one card at Q4_K_M, including Qwen3.6, Gemma 4, Mistral Small, gpt-oss-20b, and DeepSeek-R1-Distill. It covers VRAM fit, licensing, and the job each does best.

This comparison scores four leading AI coding agents—Mistral Vibe for Code, Claude Code, Cursor, and OpenAI Codex—on a real scaffold-to-PR workflow. Mistral Vibe leads with 22/25, driven by low cost, open weights, and self-hosting options. Claude Code and Codex tie at 21/25, while Cursor scores 16/25. The article details each tool's strengths and weaknesses across five dimensions: feature scaffolding, test generation, PR/async workflow, surface coverage, and cost/openness.

Mistral AI introduces a vision model that enables robots to navigate unknown environments using only a single RGB camera and natural language instructions.

Mistral AI introduced Robostral Navigate, an 8B embodied navigation model. It moves robots from a plain-language instruction using only a single RGB camera, with no LiDAR or depth sensors. The model reaches 76.6% success on R2R-CE validation unseen through a pointing method, prefix-caching training, and CISPO online reinforcement learning.

This paper explores machine learning for automatic thematic indexing of large literary corpora, using Voltaire's works as a test case. The best model, a 4-bit quantized Mistral, achieves F1 scores up to 0.67, highlighting the potential of automated indexing.

A personal, non-benchmark tier list of AI models for coding and auditing as of mid-2026, covering Anthropic Fable, OpenAI Sol, Mistral, Gemini, and DeepSeek, with commentary on US export controls and European perspectives.

AnyFile Translator is an AI-powered assistant for Google Chat that translates documents, web links, and messages while preserving original formatting. It supports over 100 languages, offers AI content writing, and ensures data privacy with encryption and deletion.

This post walks you through building a production-ready ecommerce MCP server using Amazon Bedrock AgentCore and Mistral AI Studio. It covers MCP tool implementation, two-layer JWT authentication, AWS CDK deployment, integration with Mistral AI's Vibe, and best practices for data and identity management with DynamoDB and Cognito.

This paper presents a workload-aware benchmark of KV-cache optimization techniques including KIVI, TurboQuant, SnapKV, and CaM, evaluated on Llama-3.1-8B-Instruct and Mistral-7B-Instruct-v0.3 models across multi-document QA, single-document QA, few-shot learning, and summarization tasks. Results show that compression ratio alone is a poor predictor of end-to-end performance. KIVI4 offers the most stable quality across models, SnapKV delivers the strongest long-context throughput, and CaM yields large gains on selected QA workloads but exhibits substantial workload sensitivity. The study motivates workload-aware selection of KV-cache mechanisms.