The Archive

Mistral says Le Chonk can rival top closed models while remaining open-weight.

llm-mistral 0.16 adds reasoning model support for Mistral Large 4.

Mistral releases Large 4, a 1T-param model (49B active) trained on 3,800 NVIDIA Grace Blackwell GPUs; preview via API, open weights promised end-of-month.

Simon Willison comments on Mistral Large 4 benchmarking saturation with a humorous comparative prompt test across frontier models.

French AI lab Mistral AI has released Mistral Large 4, a new large multimodal model aiming to leapfrog both American and Chinese rivals.

Empirical study of information retention in sliding-window KV inference across Qwen, Llama, Mistral models reveals cross-window state transmission.

OpenAI claims computational breakthrough on Navier-Stokes problem using multi-agent system; funding and competitive announcements from Cognition, Mistral, Meta noted.

The French AI lab has raised €3 billion at a €21 billion valuation in a Series D round led by Samsung, Scaleup Europe and PSG Equity.

IRWOZ 2.0 dataset: 390 LLM-enhanced dialogue annotations (Mistral, Claude-3.5) for industrial robot conversations across 4 domains with improved quality.

Comparative analysis of worldbuilding strategies in AI-generated vs. human fiction across GPT-4, LLaMA 3.3, Mistral 3.2, Gemma 3.

We present the REAP system for the AKBC Shared Task 2026 on constructing knowledge bases from language models in a closed-book setting, subject to a budget of at most 32B parameters and no model fine-tuning. Our system combines structured chain-of-thought reasoning, relation-specific query strategies, and a reasoning-based empty-set gate to elicit parametric knowledge, followed by direct extraction into valid JSON arrays. On the test set, the system, built on the Mistral-Small-24B-Instruct-2501 model, achieves a macro-F1 score of 0.62, with particularly strong results on countryLandBordersCou...

Quantization precision (FP16/INT8/INT4) and prompt design critically affect biomedical LLM classifier calibration on Mistral variants.

Employees of OpenAI and Anthropic, as well as Google, Meta, Thinking Machines, Microsoft, Mistral, and other leading AI labs, have written a statement to the US government supporting a potential slowdown of sorts for frontier AI development - or at least a speed-up of global coordinated governance efforts. "Al could help create a dramatically better future, but that outcome is not guaranteed," the employees wrote in a statement. "The world's leading Al companies believe they could be close to automating Al research. It is hard to predict exactly how much this will accelerate Al progress, but ...

AI companies including Nvidia and Mistral urge policymakers to avoid broad restrictions on open-weight AI models as Washington debates responses to Chinese AI and alleged model distillation.

Mistral AI, which offers some open source AI models, has raised significant funding since its creation in 2023, with the ambition to “put frontier AI in the hands of everyone.”

UKRI metascience project compares GPT-4o, Mistral, and DSIT-Taxonomies for research entity extraction from 42 funding proposal abstracts.

We talk about how this legendary investor went from humble beginnings in Singapore to leading rounds in Anthropic, Mistral, Black Forest Labs, and Periodic Labs... and the AMP secret master plan!

Recent advances in large language models (LLMs) have produced many specialized multimodal LLMs (MLLMs) that share common foundational LLMs, forming distinct model lineages. It remains unclear whether a fundamental behavioral link exists between the foundational LLMs and downstream variants. We investigate this question by quantifying head-level context-truthfulness scores. Across diverse LLM and MLLM lineages, including Vicuna-, Qwen2.5-, LLaMA2-, and Mistral-based models, we find that Truth Scores are strongly preserved within model families, even after instruction tuning or multimodal adapt...

The funding round would value the company at around €20 billion (about $23.15 billion), nearly double its Series C valuation of €11.7 billion.

This study investigates cross-lingual distributional skew (the Shibboleth Effect) in frontier large language models (LLMs) subjected to sustained adversarial conditions. We develop a multi-agent geopolitical wargame, the Cerulean Sea Crisis, a synthetic maritime territorial dispute designed to mirror the structural dynamics of Eastern Mediterranean conflicts. Six frontier models (GPT-4o, Llama-4, Mistral-Large, Gemini-3.1-Pro, Qwen3.6-Plus, and DeepSeek-R1) participate in a between-groups experiment (N = 10 games per arm, K = 5 rounds per game) in which the sole manipulation is the language o...

User reports delegating Claude Code tasks to Mistral/DeepSeek via vibe-skill tool, achieving 90% cost savings over 10 days while maintaining output quality.

Mistral AI founder tells French Parliament that engineers now manage AI agents writing code instead of writing it themselves, marking a shift in developer workflows.

Empirical study across 288 model calls identifying JSON output failures in Llama 3, Mistral, Command R, DeepSeek, Qwen; failure modes consistent across open and closed models but vary by rate.

Fairness audit of five LLMs (Gemini, GPT-4, DeepSeek, Mistral, Nemotron) on emergency triage reveals gender bias persistence in clinical decision support.

Unsloth and Mistral fixed YaRN parsing bug in Mistral Medium 3.5 inference; updated GGUFs released with mscale_all_dim correction.

Unsloth fixes broken GGUF quantizations of Mistral Medium 3.5 128B, resolving long-context degradation issues.

Mistral releases Medium 3.5, an open-weights model emphasizing reliability and robustness for production deployment.

Reddit discussion praising dense model architectures, expresses preference for continued dense model releases.

Mistral Medium 3.5 launched with modified MIT license restricting commercial use without paid license.

Mistral Medium 3.5 128B model released on Hugging Face.