Artificial intelligence,
decoded in real time.
News, comparisons and clear explanations of every AI model, updated every day. From beginner to expert.
Where do you start?
Three entry points depending on your profile and level.
Latest news
The releases, announcements and analyses that matter today.
Anthropic unveils Claude Opus 4.7 with extended reasoning mode
The new flagship sets new records on SWE-bench and GPQA, with a 200k token context window.
Mistral Large 3 launches via API: the European champion accelerates
123 billion parameter model focused on multilingual reasoning, with aggressive pricing.
The EU clarifies AI Act application for general-purpose models
New guidelines detail transparency and red-teaming obligations for LLMs.
The most up-to-date AI comparison
800+ models tracked: benchmarks, API pricing, context windows, use cases. Updated daily.
| Model | Vendor | MMLU | HumanEval | Context | Price $/M | OSS |
|---|---|---|---|---|---|---|
GPT-5 | OpenAI | 90.8 | 92.0 | 400k | 5 / 15 | |
Claude Opus 4.7 | Anthropic | 89.5 | 94.2 | 500k | 15 / 75 | |
Gemini 3 Pro | 89.0 | 90.5 | 2.0M | 1.25 / 5 | ||
DeepSeek R2 | DeepSeek | 89.0 | 92.5 | 128k | 0.55 / 2.19 |
The entire AI ecosystem, at your fingertips.
Automatic daily updates
Our scripts scan 15+ official sources every 6 hours.
Two reading levels
General audience or deep technical — you choose your depth.
Bilingual FR / EN
All content available in both languages, switch in one click.
Free, no intrusive ads
Funded by sponsored newsletter and optional Pro subscription.