Salesforce dit prévoir de dépenser 300 M$ en tokens chez Anthropic en 2026. Mon analyse, ce n’est pas qu’ils font bien les choses ou que c’est la preuve qu’ils innovent. C’est surtout qu’ils n’ont pas le choix. Le modèle des SaaS est plus menacé que jamais. Avec l’IA qui code
GENERATIVE AI
-
8B MoE Model Trained for Agentic Local Workflows
By
–
8B MoE (1B Activated) trained on 38 trillion tokens for local and agentic workflows
-
Higgsfield AI Plugin for Premiere Pro Timeline
By
–
Le plugin Higgsfield dans Premiere Pro est vraiment clean.
— Jouhatsu | AI Influence Operator (@Jouhatsu_ai) 29 mai 2026
Tout se fait directement dans la timeline :
reframe, effets AI, suppression de fond… sans quitter Adobe. https://t.co/RugFuX8lV6 pic.twitter.com/jE8PJ5hYm0Le plugin Higgsfield dans Premiere Pro est vraiment clean. Tout se fait directement dans la timeline :
reframe, effets AI, suppression de fond… sans quitter Adobe. -
ChatLLM Smart Model Router for Task-Specific AI Selection
By
–
ChatLLM smartly routes to the best model based on your prompt Opus 4.8 – front-end
GPT 5.5 – back-end coding
Grok 4.5 – real-time
Flash 3.5 – chat Nano Banana Pro – image
Seedance 2.0 – video
ElevenLabs – voice
Gemini Pro 3.1 – research
DeepSeek Flash – simple We smartly -
Visualization of all humans generated by Opus 4.8 in Claude Code
By
–
How lucky are you to have been born when and where you are? Had Opus 4.8 in Claude Code whip up a new visualization of all humans who ever lived. In addition to being neat, it is an interesting test of combining research, code, design and stats for an AI. https://
veil-of-history.netlify.app -
GPT-5 Pro models leading single-shot performance
By
–
Interesting that the GPT-5 Pro series models have consistently been the best models for single-shot attempts at the hardest problems since last summer. There has been no real competition in all that time.
-
Step 3.7 Flash MoE Model with 256K Context Released
By
–
Step 3.7 Flash is here ICYMI: 198B MoE with 11B active params, 256K context, native image + video support. Day 0 support is live on http://
build.nvidia.com with GPU-accelerated endpoints, deploy with NVIDIA NIM inference microservices, and fine-tune with the NVIDIA NeMo -
ElevenLabs Dubbing v2 Alpha: Multilingual Speech Translation
By
–
ElevenLabs introduced a new Dubbing v2 Alpha model that can translate speech across all languages while preserving the emotional tone of the original content.
— 🚨 AI News | TestingCatalog (@testingcatalog) 28 mai 2026
Big for creators 👀 https://t.co/W7dTh3IxAT pic.twitter.com/gnC1FLeG4kElevenLabs introduced a new Dubbing v2 Alpha model that can translate speech across all languages while preserving the emotional tone of the original content. Big for creators
-
AI Reasoning Laziness and Prompt Injection Solutions
By
–
Anthropic semble avoir trouvé une solution aux problèmes où l’IA vous donne une réponse incorrect car elle a la flemme et préfère arrêter le raisonnement. Moi je dis ils doivent faire du prompt injections à chaque thinking avec des menaces .
-
Anthropic Acceleration: Model Release Cycle Shortening Trend
By
–
Anthropic released the next version sooner than I thought – the trend is accelerating – from 50-70 days before, down to 42 days since Opus 4.7