https://www.askthisguy.com/fr/ 2026-11-10 https://www.askthisguy.com/en/ 2026-11-10 https://www.askthisguy.com/fr/blog/ 2026-11-10 https://www.askthisguy.com/en/blog/ 2026-11-10 https://www.askthisguy.com/fr/demo/ 2026-11-10 https://www.askthisguy.com/en/demo/ 2026-11-10 https://www.askthisguy.com/fr/platform/ 2026-11-10 https://www.askthisguy.com/en/platform/ 2026-11-10 https://www.askthisguy.com/fr/ia-souveraine/ 2026-11-10 https://www.askthisguy.com/en/sovereign-ai/ 2026-11-10 https://www.askthisguy.com/fr/about/ 2026-11-10 https://www.askthisguy.com/en/about/ 2026-11-10 https://www.askthisguy.com/fr/doc/ 2026-11-10 https://www.askthisguy.com/en/doc/ 2026-11-10 https://www.askthisguy.com/fr/author/jean-christophe-budin/ 2026-11-10 https://www.askthisguy.com/fr/author/remi-chkaibane/ 2026-11-09 https://www.askthisguy.com/en/author/jean-christophe-budin/ 2026-11-10 https://www.askthisguy.com/en/author/remi-chkaibane/ 2026-11-09 https://www.askthisguy.com/fr/doc/admin-console/knowledge-management/connectors/conf-guides/ 2026-10-16 https://www.askthisguy.com/fr/doc/admin-console/organization-settings/ 2026-11-06 https://www.askthisguy.com/fr/doc/customizing-atg-to-you/introduction/ 2026-06-22 https://www.askthisguy.com/fr/doc/admin-console/knowledge-management/ 2026-11-06 https://www.askthisguy.com/fr/doc/atg-embedded/how-to-integrate-atg-embedded-in-your-webapp/ 2026-11-06 https://www.askthisguy.com/fr/doc/customizing-atg-to-you/tailored-processing/ 2026-11-06 https://www.askthisguy.com/fr/doc/admin-console/tools-management/ 2026-11-06 https://www.askthisguy.com/fr/doc/atg-embedded/install-on-wordpress-with-the-atg-plugin/ 2026-11-06 https://www.askthisguy.com/fr/doc/customizing-atg-to-you/tailored-tools/ 2026-05-05 https://www.askthisguy.com/fr/doc/admin-console/tools-management/chat-tools/ 2026-11-06 https://www.askthisguy.com/fr/doc/admin-console/tools-management/mcp-servers/ 2026-11-06 https://www.askthisguy.com/fr/doc/admin-console/tools-management/mcp-catalog/ 2026-11-06 https://www.askthisguy.com/fr/doc/admin-console/how-to-configure-the-agents/ 2026-11-06 https://www.askthisguy.com/fr/doc/customizing-atg-to-you/database-query/ 2026-11-06 https://www.askthisguy.com/fr/doc/admin-console/voice-mode/ 2026-11-06 https://www.askthisguy.com/fr/doc/admin-console/agent-owners/ 2026-11-06 https://www.askthisguy.com/fr/doc/admin-console/tags-alerts/ 2026-11-06 https://www.askthisguy.com/fr/doc/admin-console/agent-evaluation/ 2026-06-22 https://www.askthisguy.com/fr/doc/admin-console/improvement-tracker/ 2026-11-06 https://www.askthisguy.com/fr/doc/admin-console/conversations/ 2026-11-06 https://www.askthisguy.com/fr/doc/admin-console/insights/ 2026-11-06 https://www.askthisguy.com/fr/doc/admin-console/knowledge-management/connectors/conf-guides/confluence-conf-guide/ 2026-11-06 https://www.askthisguy.com/fr/doc/admin-console/knowledge-management/connectors/confluence-connector/ 2026-11-06 https://www.askthisguy.com/fr/doc/admin-console/users-management/ 2026-11-06 https://www.askthisguy.com/fr/doc/admin-console/knowledge-management/connectors/conf-guides/google-drive-conf-guide/ 2026-11-06 https://www.askthisguy.com/fr/doc/admin-console/knowledge-management/connectors/ 2026-11-06 https://www.askthisguy.com/fr/doc/admin-console/ 2026-11-06 https://www.askthisguy.com/fr/doc/admin-console/knowledge-management/connectors/google-drive-connector/ 2026-11-06 https://www.askthisguy.com/fr/doc/admin-console/knowledge-management/file-upload/ 2026-11-06 https://www.askthisguy.com/fr/doc/admin-console/knowledge-management/connectors/conf-guides/notion-conf-guide/ 2026-11-06 https://www.askthisguy.com/fr/doc/admin-console/knowledge-management/connectors/notion-connector/ 2026-11-06 https://www.askthisguy.com/fr/doc/admin-console/knowledge-management/how-is-my-companys-data-processed-what-is-rag/ 2026-11-06 https://www.askthisguy.com/fr/doc/admin-console/knowledge-management/connectors/conf-guides/sharepoint-onedrive-conf-guide/ 2026-11-06 https://www.askthisguy.com/fr/doc/admin-console/knowledge-management/how-documents-images-are-processed-by-atg/ 2026-06-22 https://www.askthisguy.com/fr/doc/admin-console/knowledge-management/connectors/sharepoint-onedrive-connector/ 2026-11-06 https://www.askthisguy.com/fr/doc/admin-console/knowledge-management/connectors/conf-guides/slack-conf-guide/ 2026-11-06 https://www.askthisguy.com/fr/doc/admin-console/knowledge-management/connectors/ftp-connector/ 2026-11-06 https://www.askthisguy.com/fr/doc/admin-console/knowledge-management/connectors/slack-connector/ 2026-11-06 https://www.askthisguy.com/fr/doc/admin-console/knowledge-management/connectors/web-sites-connector/ 2026-11-06 https://www.askthisguy.com/fr/doc/atg-embedded/ 2026-11-06 https://www.askthisguy.com/fr/doc/admin-console/organization-settings/choosing-the-right-ai-policy/ 2026-10-21 https://www.askthisguy.com/fr/doc/admin-console/organization-settings/predefined-prompts/ 2026-08-03 https://www.askthisguy.com/fr/doc/webapp/ 2026-11-06 https://www.askthisguy.com/fr/doc/webapp/using-voice-mode/ 2026-11-06 https://www.askthisguy.com/fr/doc/webapp/generating-images/ 2026-11-06 https://www.askthisguy.com/fr/doc/webapp/llm-best-practices/ 2026-03-26 https://www.askthisguy.com/fr/doc/webapp/using-attachments/ 2026-08-03 https://www.askthisguy.com/fr/doc/infrastructure-and-security/ai-providers/ 2026-11-06 https://www.askthisguy.com/fr/doc/infrastructure-and-security/data-management/ 2026-11-06 https://www.askthisguy.com/fr/doc/infrastructure-and-security/sovereign-ai/ 2026-11-06 https://www.askthisguy.com/fr/doc/customizing-atg-to-you/ 2026-11-06 https://www.askthisguy.com/fr/doc/infrastructure-and-security/ 2026-11-06 https://www.askthisguy.com/fr/doc/frequently-asked-questions/ 2026-11-10 https://www.askthisguy.com/en/doc/admin-console/knowledge-management/connectors/conf-guides/ 2026-10-16 https://www.askthisguy.com/en/doc/admin-console/organization-settings/ 2026-11-06 https://www.askthisguy.com/en/doc/customizing-atg-to-you/introduction/ 2026-06-22 https://www.askthisguy.com/en/doc/admin-console/knowledge-management/ 2026-11-06 https://www.askthisguy.com/en/doc/atg-embedded/how-to-integrate-atg-embedded-in-your-webapp/ 2026-11-06 https://www.askthisguy.com/en/doc/customizing-atg-to-you/tailored-processing/ 2026-11-06 https://www.askthisguy.com/en/doc/admin-console/tools-management/ 2026-11-06 https://www.askthisguy.com/en/doc/atg-embedded/install-on-wordpress-with-the-atg-plugin/ 2026-11-06 https://www.askthisguy.com/en/doc/customizing-atg-to-you/tailored-tools/ 2026-05-05 https://www.askthisguy.com/en/doc/admin-console/tools-management/chat-tools/ 2026-11-06 https://www.askthisguy.com/en/doc/admin-console/tools-management/mcp-servers/ 2026-11-06 https://www.askthisguy.com/en/doc/admin-console/tools-management/mcp-catalog/ 2026-11-06 https://www.askthisguy.com/en/doc/admin-console/how-to-configure-the-agents/ 2026-11-06 https://www.askthisguy.com/en/doc/customizing-atg-to-you/database-query/ 2026-11-06 https://www.askthisguy.com/en/doc/admin-console/voice-mode/ 2026-11-06 https://www.askthisguy.com/en/doc/admin-console/agent-owners/ 2026-11-06 https://www.askthisguy.com/en/doc/admin-console/tags-alerts/ 2026-11-06 https://www.askthisguy.com/en/doc/admin-console/agent-evaluation/ 2026-06-22 https://www.askthisguy.com/en/doc/admin-console/improvement-tracker/ 2026-11-06 https://www.askthisguy.com/en/doc/admin-console/conversations/ 2026-11-06 https://www.askthisguy.com/en/doc/admin-console/insights/ 2026-11-06 https://www.askthisguy.com/en/doc/admin-console/knowledge-management/connectors/conf-guides/confluence-conf-guide/ 2026-11-06 https://www.askthisguy.com/en/doc/admin-console/knowledge-management/connectors/confluence-connector/ 2026-11-06 https://www.askthisguy.com/en/doc/admin-console/users-management/ 2026-11-06 https://www.askthisguy.com/en/doc/admin-console/knowledge-management/connectors/conf-guides/google-drive-conf-guide/ 2026-11-06 https://www.askthisguy.com/en/doc/admin-console/knowledge-management/connectors/ 2026-11-06 https://www.askthisguy.com/en/doc/admin-console/ 2026-11-06 https://www.askthisguy.com/en/doc/admin-console/knowledge-management/connectors/google-drive-connector/ 2026-11-06 https://www.askthisguy.com/en/doc/admin-console/knowledge-management/file-upload/ 2026-11-06 https://www.askthisguy.com/en/doc/admin-console/knowledge-management/connectors/conf-guides/notion-conf-guide/ 2026-11-06 https://www.askthisguy.com/en/doc/admin-console/knowledge-management/connectors/notion-connector/ 2026-11-06 https://www.askthisguy.com/en/doc/admin-console/knowledge-management/how-is-my-companys-data-processed-what-is-rag/ 2026-11-06 https://www.askthisguy.com/en/doc/admin-console/knowledge-management/connectors/conf-guides/sharepoint-onedrive-conf-guide/ 2026-11-06 https://www.askthisguy.com/en/doc/admin-console/knowledge-management/how-documents-images-are-processed-by-atg/ 2026-06-22 https://www.askthisguy.com/en/doc/admin-console/knowledge-management/connectors/sharepoint-onedrive-connector/ 2026-11-06 https://www.askthisguy.com/en/doc/admin-console/knowledge-management/connectors/conf-guides/slack-conf-guide/ 2026-11-06 https://www.askthisguy.com/en/doc/admin-console/knowledge-management/connectors/ftp-connector/ 2026-11-06 https://www.askthisguy.com/en/doc/admin-console/knowledge-management/connectors/slack-connector/ 2026-11-06 https://www.askthisguy.com/en/doc/admin-console/knowledge-management/connectors/web-sites-connector/ 2026-11-06 https://www.askthisguy.com/en/doc/atg-embedded/ 2026-11-06 https://www.askthisguy.com/en/doc/admin-console/organization-settings/choosing-the-right-ai-policy/ 2026-10-21 https://www.askthisguy.com/en/doc/admin-console/organization-settings/predefined-prompts/ 2026-08-03 https://www.askthisguy.com/en/doc/webapp/ 2026-11-06 https://www.askthisguy.com/en/doc/webapp/using-voice-mode/ 2026-11-06 https://www.askthisguy.com/en/doc/webapp/generating-images/ 2026-11-06 https://www.askthisguy.com/en/doc/webapp/llm-best-practices/ 2026-03-26 https://www.askthisguy.com/en/doc/webapp/using-attachments/ 2026-08-03 https://www.askthisguy.com/en/doc/infrastructure-and-security/ai-providers/ 2026-11-06 https://www.askthisguy.com/en/doc/infrastructure-and-security/data-management/ 2026-11-06 https://www.askthisguy.com/en/doc/infrastructure-and-security/sovereign-ai/ 2026-11-06 https://www.askthisguy.com/en/doc/customizing-atg-to-you/ 2026-11-06 https://www.askthisguy.com/en/doc/infrastructure-and-security/ 2026-11-06 https://www.askthisguy.com/en/doc/frequently-asked-questions/ 2026-11-10 https://www.askthisguy.com/fr/blog/qwen-4-vs-qwen-3-8-27b/ 2026-11-10 https://www.askthisguy.com/blog/qwen-4-vs-qwen-3-8-27b/amd-qwen-3-8-27b-ryzen-radeon.webp https://www.askthisguy.com/blog/qwen-4-vs-qwen-3-8-27b/rythme-sorties-qwen-2026-fr.webp Frise des sorties Qwen en 2026, sur quatre lignes. 27B dense à poids ouverts : Qwen 3.5 27B le 24 février, Qwen 3.6 27B le 22 avril (deux mois plus tard), Qwen 3.8 27B le 14 août (près de quatre mois plus tard). Autres modèles ouverts : Qwen 3.5 397B le 16 février, Qwen 3.6 35B-A3B le 16 avril, Qwen 3.8 2.4T sous licence maison le 12 août, Qwen 3.8 Flash-Next, aperçu de Qwen 4, le 26 août. Modèles fermés : Qwen 3.7 Max le 20 mai, Qwen 3.7 Plus le 1er juin, Qwen 3.8 Max le 3 août. Qwen 4 : annoncé en entraînement le 22 septembre, « Qwen4 next week » rapporté le 29 octobre, teaser officiel le 6 novembre https://www.askthisguy.com/blog/qwen-4-vs-qwen-3-8-27b/benchmarks-qwen-27b-fr.webp Graphique en barres des scores publiés par Qwen, sur 100, pour Qwen 3.6 27B, Qwen 3.8 27B et Opus 4.6 Max : SWE-bench Pro 53,5, 61,7 et 53,4 ; Terminal Bench 2.1 63,4, 73,0 et 78,2 ; OSWorld-Verified 63,9, 84,3 et 72,7 ; IFBench 69,1, 79,5 et 62,5 ; GPQA Diamond 87,8, 89,2 et 91,3 ; HLE 24,0, 30,8 et 40,0 https://www.askthisguy.com/blog/qwen-4-vs-qwen-3-8-27b/poids-vs-intelligence-fr.webp Nuage de points, paramètres totaux en échelle logarithmique et indice d'Artificial Analysis v5.0.1. Qwen 3.8 27B, dense, 27 milliards de paramètres, note 20, seul modèle dans la zone qui tient sur un GPU de 80 Go. Mistral Small 4, MoE de 119 milliards, note 4. Mistral Medium 3.5, dense de 128 milliards, note 5. Mistral Large 4, MoE d'environ 1 000 milliards, note 26. Qwen 3.8 2.4T, MoE de 2 400 milliards, placé à 34, la note de sa version API Qwen 3.8 Max. Une flèche indique que le meilleur est en haut à gauche : moins de paramètres, meilleure note https://www.askthisguy.com/fr/blog/ia-open-source-licences/ 2026-11-10 https://www.askthisguy.com/blog/ia-open-source-licences/giammarco-boscaro-zeh-ljawhtg-unsplash.webp https://www.askthisguy.com/fr/blog/text-to-sql-agent/ 2026-11-09 https://www.askthisguy.com/blog/text-to-sql-agent/multi-database-sql-agent.webp https://www.askthisguy.com/fr/blog/reranking-rag/ 2026-11-06 https://www.askthisguy.com/blog/reranking-rag/pierre-bamin-lm4-nmcj4xk-unsplash.webp https://www.askthisguy.com/fr/blog/choisis-un-nombre-entre-1-et-10/ 2026-11-05 https://www.askthisguy.com/blog/choisis-un-nombre-entre-1-et-10/riho-kroll-m4sgyahyn5o-unsplash.webp https://www.askthisguy.com/blog/choisis-un-nombre-entre-1-et-10/part-du-7-par-modele-fr.webp Part des réponses « 7 » par modèle à température 1, en français et en anglais : 100 % pour Claude Haiku 5.5, GPT-6 Luna, Gemini 3.8 Flash et Gemma 4 31B, 96 à 100 % pour Qwen 3.8 27B, 74 à 76 % pour Mistral Large 4 et 65 à 70 % pour Mistral Small ; les humains sont à 33 % et le hasard à 10 % https://www.askthisguy.com/blog/choisis-un-nombre-entre-1-et-10/temperature-part-du-7-fr.webp Part des réponses « 7 » selon la température, en français : Gemini 3.8 Flash, GPT-6 Luna et Gemma 4 31B restent à 100 % de 0 à 1,5 ; Qwen 3.8 27B descend à environ 95 % à 1,5 ; Mistral Large 4 reste entre 71 et 81 % ; Mistral Small descend de 100 % à environ 40 % ; le hasard est à 10 % https://www.askthisguy.com/fr/blog/audit-rag/ 2026-11-04 https://www.askthisguy.com/blog/audit-rag/berke-citak-qv5voi18ld8-unsplash.webp https://www.askthisguy.com/fr/blog/forward-deployed-engineer-definition/ 2026-11-03 https://www.askthisguy.com/blog/forward-deployed-engineer-definition/vitaly-gariev-cn54mbf1f8i-unsplash.webp https://www.askthisguy.com/fr/blog/mixture-of-experts-dense-vs-moe/ 2026-10-28 https://www.askthisguy.com/blog/mixture-of-experts-dense-vs-moe/fotoflo-iq3c6trhsow-unsplash.webp https://www.askthisguy.com/fr/blog/data-room-ia-mcp/ 2026-10-27 https://www.askthisguy.com/blog/data-room-ia-mcp/cover.webp https://www.askthisguy.com/fr/blog/ia-due-diligence-trouver-le-bon-document/ 2026-10-23 https://www.askthisguy.com/blog/ia-due-diligence-trouver-le-bon-document/couverture.webp https://www.askthisguy.com/blog/ia-due-diligence-trouver-le-bon-document/trouver-le-document-escalier-fr.webp Graphique en barres : GPT-4 Turbo répond juste à 9 % des questions sans document, 19 % quand il cherche dans toute la base de rapports, 50 % avec le bon rapport en extraits, 79 % avec le bon rapport en entier, 85 % avec la bonne page https://www.askthisguy.com/blog/ia-due-diligence-trouver-le-bon-document/logiciel-vs-modele-fr.webp Graphique en deux parties. À gauche, même modèle et logiciel différent : Mistral Medium 3.5 passe de 22,7 % à 70 % puis 78,7 %, GLM-5.2 de 26,7 % à 79,3 % puis 86 %. À droite, même logiciel et modèle différent chez Dewey : 62,9 % pour GPT-5.4, 83,7 % pour Claude Opus 4.6 https://www.askthisguy.com/fr/blog/chatgpt-enterprise-credits/ 2026-11-06 https://www.askthisguy.com/blog/chatgpt-enterprise-credits/portefeuille-vide.webp https://www.askthisguy.com/blog/chatgpt-enterprise-credits/prix-intelligence-modeles-fr.webp Deux graphiques en barres alignés par modèle. À gauche, le prix pour les mêmes tokens : GPT-6 Luna 1, GPT-5.6 Luna 2, GPT-6.1 Sol 20, GPT-6 Sol 20, GPT-5.6 Terra 20 à 24, GPT-5.6 Sol 40, GPT-6 Astra 100. À droite, le score d'intelligence : 38, 37, 52, 48, 42, 47 et 53. GPT-6 Luna et GPT-6.1 Sol sont mis en évidence comme recommandés, GPT-6 Astra comme exceptionnel, les autres en gris https://www.askthisguy.com/blog/chatgpt-enterprise-credits/changer-modele-chatgpt-excel-fr.webp Schéma en trois étapes du panneau ChatGPT dans Excel : cliquer sur le nom du modèle « GPT-5.6 Sol » en bas du panneau, choisir « GPT-6 Luna » dans la liste des modèles, puis garder l'effort de réflexion sur « Moyen » ou moins https://www.askthisguy.com/fr/blog/ovhcloud-strategie-ia/ 2026-10-20 https://www.askthisguy.com/blog/ovhcloud-strategie-ia/octave-klaba-ovhcloud-datacenter.webp https://www.askthisguy.com/blog/ovhcloud-strategie-ia/prix-tokens-europe-fr.webp Prix par million de tokens en euros HT au 20 octobre 2026, en entrée puis en sortie. gpt-oss-120b : OVHcloud 0,08 et 0,40, Scaleway 0,15 et 0,60, IONOS 0,15 et 0,65. Llama 3.3 70B : OVHcloud 0,67 et 0,67, Scaleway 0,90 et 0,90, IONOS 0,65 et 0,65. Mistral Small 3.2 24B : OVHcloud 0,09 et 0,28, Scaleway 0,15 et 0,35, IONOS 0,10 et 0,30. Qwen3.5-397B : 0,60 et 3,60 chez les trois https://www.askthisguy.com/blog/ovhcloud-strategie-ia/prix-ovhcloud-umans-fr.webp OVHcloud face à Umans au 20 octobre 2026, prix par million de tokens en euros et indice d'intelligence d'Artificial Analysis. Entrée de gamme : gpt-oss-120b chez OVHcloud, 0,08 € en entrée, 0,40 € en sortie, indice 12 ; DeepSeek V4 Flash chez Umans, 0,12 €, 0,25 €, indice 34. Haut de gamme : Qwen3.5-397B chez OVHcloud, 0,60 €, 3,60 €, indice 18 ; GLM 5.3 chez Umans, 1,25 €, 3,92 €, indice 45. Prix de Umans convertis depuis le dollar au cours de la BCE du 19 octobre 2026 https://www.askthisguy.com/fr/blog/timestral-bitcoin-prevision/ 2026-10-19 https://www.askthisguy.com/blog/timestral-bitcoin-prevision/beatriz-cattel-c55m3wxaob0-unsplash.webp https://www.askthisguy.com/blog/timestral-bitcoin-prevision/prix-volume-par-annee-fr.webp Gain de Timestral-209M année par année : sur le prix à une heure, de +0,57 % en 2019 à -0,05 % en 2026 ; sur le volume du jour suivant, toujours positif, +6,3 % en 2026 https://www.askthisguy.com/fr/blog/claude-mythos-failles-correctifs/ 2026-10-16 https://www.askthisguy.com/blog/claude-mythos-failles-correctifs/mythos-failles-correctifs.webp https://www.askthisguy.com/fr/blog/coreweave-dette-dependance-ia/ 2026-10-16 https://www.askthisguy.com/blog/coreweave-dette-dependance-ia/m-w7weqy01k3s-unsplash.webp https://www.askthisguy.com/fr/blog/french-tech-ownership/ 2026-11-06 https://www.askthisguy.com/blog/french-tech-ownership/fr-to-us.webp https://www.askthisguy.com/fr/blog/ai-self-replication/ 2026-10-14 https://www.askthisguy.com/blog/ai-self-replication/ai-worms.webp https://www.askthisguy.com/fr/blog/anthropic-scion-git/ 2026-11-06 https://www.askthisguy.com/blog/anthropic-scion-git/de-git-a-scion-migration-moderne.webp https://www.askthisguy.com/fr/blog/ai-distillation/ 2026-10-13 https://www.askthisguy.com/blog/ai-distillation/francois-hoppet-kqohccwt7tq-unsplash.webp https://www.askthisguy.com/fr/blog/llm-gateway/ 2026-11-06 https://www.askthisguy.com/blog/llm-gateway/denys-nevozhai-7nrsvjvalna-unsplash.webp https://www.askthisguy.com/blog/enterprise-ai-sla/success-rate-per-provider.webp Taux de succès par fournisseur d'inférence : Scaleway 99,87 %, Nebius 99,81 %, Cerebras 99,78 %, DeepInfra 99,76 %, Mistral 99,56 % https://www.askthisguy.com/blog/llm-gateway/meme-modele-debit-prix-par-fournisseur.webp Deux graphiques en barres. À gauche, le modèle gpt-oss-120b : Nebius le sert à 40 tokens par seconde et Groq à 500, tous deux à 0,15 $ en entrée et 0,60 $ en sortie par million de tokens, Cerebras à 3 000 tokens par seconde pour 0,35 et 0,75 $. À droite, le modèle DeepSeek V4 Pro coûte 0,57 $ par million de tokens chez DeepSeek (Chine), 1,69 $ chez DeepInfra (États-Unis), 2,11 $ chez Umans et 2,21 $ chez Arkane Cloud (France), 2,27 $ chez Nebius (Pays-Bas). https://www.askthisguy.com/fr/blog/rag-how-many-documents/ 2026-10-07 https://www.askthisguy.com/blog/rag-how-many-documents/schema-rag-filtrer-avant-de-chercher.svg https://www.askthisguy.com/fr/blog/chatgpt-watermark/ 2026-10-15 https://www.askthisguy.com/blog/chatgpt-watermark/code-ecran-sombre.webp https://www.askthisguy.com/fr/blog/mistral-ocr-paddleocr-cost/ 2026-09-30 https://www.askthisguy.com/blog/mistral-ocr-paddleocr-cost/viktor-talashuk-05hlfqu8bfw-unsplash.webp https://www.askthisguy.com/blog/mistral-ocr-paddleocr-cost/cout-ocr-million-pages.svg Diagramme en barres comparant le coût de l'OCR d'un million de pages : 4 000 $ avec Mistral OCR 4.1, 2 000 $ avec Mistral OCR 3 ou OCR 4 en batch, 151 $ avec PaddleOCR-VL sur un GPU RTX PRO 6000 à la demande chez Verda, 76 $ sur le même GPU en tarif spot https://www.askthisguy.com/fr/blog/jev-rag-reranking/ 2026-10-15 https://www.askthisguy.com/blog/jev-rag-reranking/daniel-forsman-ph4zjrwf4x8-unsplash.webp https://www.askthisguy.com/fr/blog/how-ai-works-key-concepts/ 2026-09-07 https://www.askthisguy.com/blog/how-ai-works-key-concepts/markus-spiske-vic0envgdtu-unsplash.webp https://www.askthisguy.com/blog/how-ai-works-key-concepts/schema-reseau-de-neurones.webp Schéma d'un réseau de neurones : une couche d'entrée reçoit des photos de chats et de chiens, les signaux traversent plusieurs couches internes de neurones interconnectés, et la couche de sortie sépare les images en deux catégories, chats et chiens https://www.askthisguy.com/blog/how-ai-works-key-concepts/intelligence-vs-parametres-modeles.webp Nuage de points comparant le niveau des modèles ouverts à leur nombre de paramètres, en échelle logarithmique. La courbe monte fortement jusqu'à une trentaine de milliards de paramètres, puis s'aplatit : au-delà, multiplier la taille par cent ne fait plus gagner que quelques points https://www.askthisguy.com/blog/how-ai-works-key-concepts/intelligence-index-modeles-par-pays.webp Classement des modèles selon un indice de niveau général, avec le drapeau du pays d'origine. Les deux premiers sont américains, les deux suivants chinois et très proches, les modèles français arrivent nettement derrière et le modèle suisse ferme la marche https://www.askthisguy.com/blog/how-ai-works-key-concepts/tokens-francais-anglais-swahili.webp Le même texte découpé en tokens dans trois langues. En français, 2 658 caractères donnent 692 tokens. En anglais, 2 589 caractères n'en donnent que 606. En swahili, 2 688 caractères en produisent 835 https://www.askthisguy.com/blog/how-ai-works-key-concepts/vitesse-fournisseurs-inference.webp Vitesse de génération d'un même modèle ouvert selon le fournisseur qui l'exécute, avec le drapeau du pays. Les quatre premiers sont américains, le plus rapide dépassant 1 900 tokens par seconde ; les deux fournisseurs européens ferment la marche, à 243 et 159 tokens par seconde https://www.askthisguy.com/blog/how-ai-works-key-concepts/schema-appel-outil.webp Schéma d'un appel d'outil : la question de l'utilisateur arrive au modèle, qui choisit soit de répondre directement, soit de déclencher un outil exécuté par le logiciel de l'entreprise (ERP, CRM, base documentaire, recherche web, calculs, envois d'e-mails) avant de formuler sa réponse https://www.askthisguy.com/fr/blog/hybrid-search-retrieval-methods/ 2026-08-26 https://www.askthisguy.com/blog/hybrid-search-retrieval-methods/schema-methodes-recherche-complementaires.svg https://www.askthisguy.com/fr/blog/ia-francaise-souverainete/ 2026-10-15 https://www.askthisguy.com/blog/ia-francaise-souverainete/chris-karidis-nnzkZNYWHaU-unsplash.webp https://www.askthisguy.com/fr/blog/enterprise-ai-inference/ 2026-10-12 https://www.askthisguy.com/blog/enterprise-ai-inference/ilias-gainutdinov-UILFtxyE9Eg-unsplash.webp https://www.askthisguy.com/blog/enterprise-ai-inference/cout-vs-intelligence-modeles-ia.webp Nuage de points comparant coût au million de tokens et indice d'intelligence pour quatre offres. DeepSeek V4 Flash servi par le français Umans atteint un indice de 50 pour 0,11 € au MTok, quand GPT-5.6 Sol d'OpenAI atteint 56 pour 7,35 €, et GPT-5.6 Luna 46 pour 0,29 €. https://www.askthisguy.com/fr/blog/enterprise-ai-agents/ 2026-08-08 https://www.askthisguy.com/blog/enterprise-ai-agents/campaign-creators-kq4tbklji-unsplash.webp https://www.askthisguy.com/fr/blog/internal-company-chatbot/ 2026-09-07 https://www.askthisguy.com/blog/internal-company-chatbot/philipp-katzenberger-iijruoerocq-unsplash.webp https://www.askthisguy.com/blog/internal-company-chatbot/chatbot-interne-entreprise-sources.webp Schéma d'un chatbot interne entreprise : les sources (Drive, SharePoint, Confluence, Notion, intranet) alimentent un moteur RAG qui applique les droits de chacun et répond en citant l'accord RTT 2026 https://www.askthisguy.com/fr/blog/knowledge-management-why-it-fails/ 2026-08-26 https://www.askthisguy.com/blog/knowledge-management-why-it-fails/salle-etude-vide.webp https://www.askthisguy.com/fr/blog/enterprise-ai-sla/ 2026-10-12 https://www.askthisguy.com/blog/enterprise-ai-sla/bermix-studio-wbc8mlfuzu-unsplash.webp https://www.askthisguy.com/blog/enterprise-ai-sla/microsoft-availability.webp Disponibilité mondiale Microsoft 365 par trimestre, autour de 99,9 % et plus (source Microsoft Learn) https://www.askthisguy.com/blog/enterprise-ai-sla/success-rate-per-provider.webp Taux de succès par provider d'inférence : Scaleway 99,87 %, Nebius 99,81 %, Cerebras 99,78 %, DeepInfra 99,76 %, Mistral 99,56 % https://www.askthisguy.com/fr/blog/chatbot-produit-saas/ 2026-11-10 https://www.askthisguy.com/blog/chatbot-produit-saas/ant-rozetsky-r4iD__fTqIs-unsplash.webp https://www.askthisguy.com/blog/chatbot-produit-saas/atg-console-aide-sharepoint-fr.webp Widget d'aide dans la console d'administration Ask This Guy, qui guide un administrateur étape par étape pour configurer le connecteur SharePoint https://www.askthisguy.com/doc/tags-alerts-01-fr.webp Tags et alertes : trois règles configurées, intérêt fonctionnalité premium, intention d'achat et client veut annuler https://www.askthisguy.com/fr/blog/data-analysis-tools/ 2026-07-08 https://www.askthisguy.com/blog/data-analysis-tools/ux-indonesia-8mikj83lmsq-unsplash.webp https://www.askthisguy.com/blog/data-analysis-tools/illustration-excel.webp Tableur Excel avec formules, données et graphiques https://www.askthisguy.com/blog/data-analysis-tools/illustration-powerbi.webp Tableau de bord Power BI avec graphiques et indicateurs de suivi https://www.askthisguy.com/blog/data-analysis-tools/illustration-jupyter.webp Notebook Jupyter avec code Python et visualisation de données https://www.askthisguy.com/blog/data-analysis-tools/illustration-matlab.webp Interface MATLAB pour modélisation numérique et analyse statistique https://www.askthisguy.com/blog/data-analysis-tools/illustration-ask-this-guy.webp Interface Talk to Data d'Ask This Guy pour interroger ses données en langage naturel https://www.askthisguy.com/fr/blog/financial-analysis-without-excel/ 2026-10-15 https://www.askthisguy.com/blog/financial-analysis-without-excel/carlos-muza-hpjSkU2UYSU-unsplash.webp https://www.askthisguy.com/fr/blog/qualified-leads-website-chatbot/ 2026-09-07 https://www.askthisguy.com/blog/qualified-leads-website-chatbot/juan-rodriguez-duarte-31NZ7nGupm8-unsplash.webp https://www.askthisguy.com/fr/blog/contextual-ai-rag-use-cases/ 2026-05-05 https://www.askthisguy.com/blog/contextual-ai-rag-use-cases/vitaly-gariev-Vvn0OD0IxBM-unsplash.webp https://www.askthisguy.com/fr/blog/enterprise-rag-production-mistakes/ 2026-08-05 https://www.askthisguy.com/blog/enterprise-rag-production-mistakes/maksym-kaharlytskyi-Q9y3LRuuxmg-unsplash.webp https://www.askthisguy.com/blog/enterprise-rag-production-mistakes/success-rate-q1-text-generation.webp Benchmark ATG Q1 2026 des taux de succès des fournisseurs de génération de texte https://www.askthisguy.com/fr/blog/use-minimal-ai-automation/ 2026-03-26 https://www.askthisguy.com/blog/use-minimal-ai-automation/paul-lichtblau--n_bYa2Bhwg-unsplash.webp https://www.askthisguy.com/fr/blog/databases-gold-mine-ai-shovels/ 2026-03-18 https://www.askthisguy.com/blog/databases-gold-mine-ai-shovels/gold-mine.webp https://www.askthisguy.com/blog/databases-gold-mine-ai-shovels/you-dig.webp Illustration : « Dans ce monde il y a deux sortes de gens : ceux qui ont des armes chargées et ceux qui creusent. Tu creuses. » https://www.askthisguy.com/blog/databases-gold-mine-ai-shovels/claudio-schwarz-fyeOxvYvIyY-unsplash.webp Crédit : Claudio Schwarz https://www.askthisguy.com/fr/blog/llm-local-enterprise-sovereignty-illusion/ 2026-10-15 https://www.askthisguy.com/blog/llm-local-enterprise-sovereignty-illusion/data-center.webp https://www.askthisguy.com/blog/llm-local-enterprise-sovereignty-illusion/agentic-ai-dobi-2.webp IA agentique : enchaînement d'étapes et appels d'outils https://www.askthisguy.com/blog/llm-local-enterprise-sovereignty-illusion/small-is-beautiful-2.webp Quand les petits modèles suffisent https://www.askthisguy.com/fr/blog/how-mcp-accelerate-agentic-ai/ 2026-10-15 https://www.askthisguy.com/blog/how-mcp-accelerate-agentic-ai/water-connection.webp https://www.askthisguy.com/blog/how-mcp-accelerate-agentic-ai/tool-calling2.webp Flux d'appel d'un outil par l'agent IA https://www.askthisguy.com/blog/how-mcp-accelerate-agentic-ai/tool-discovery-selection.webp Processus de découverte et sélection des outils MCP https://www.askthisguy.com/blog/how-mcp-accelerate-agentic-ai/mcp-ui-flow.webp Configuration et sélection des outils MCP dans l'interface d'administration https://www.askthisguy.com/blog/how-mcp-accelerate-agentic-ai/user-surf-chat.webp L'UI de Chat d'Ask This Guy https://www.askthisguy.com/fr/blog/embedding-strategy/ 2026-01-28 https://www.askthisguy.com/blog/embedding-strategy/chuttersnap-cGXdjyP6-NU-unsplash.webp https://www.askthisguy.com/blog/embedding-strategy/mistral_openai_table.webp Performances Mistral & OpenAI https://www.askthisguy.com/blog/embedding-strategy/mistral_openai_1s.webp Performances Mistral & OpenAI https://www.askthisguy.com/blog/embedding-strategy/providers_comparison.webp Comparaison des performances de fournisseurs d'embeddings https://www.askthisguy.com/blog/embedding-strategy/european_providers_1s.webp Comparaison des performances de fournisseurs d'embeddings https://www.askthisguy.com/fr/blog/rag-tutorial/ 2026-10-12 https://www.askthisguy.com/blog/rag-tutorial/scott-graham-5fNmWej4tAA-unsplash%20(1).webp https://www.askthisguy.com/blog/rag-tutorial/ai-lost.webp IA incapable de répondre à des questions sur des documents internes d'entreprise https://www.askthisguy.com/blog/rag-tutorial/bibliothecaire.webp Un bibliothécaire rapide et précis https://www.askthisguy.com/blog/rag-tutorial/schema-rag-fonctionnement.svg Schéma RAG en deux phases : préparation des sources (analyse documentaire, découpage en chunks, indexation en vecteurs et mots-clés) puis traitement d'une question (recherche hybride, extraits pertinents, génération par le LLM, réponse sourcée) https://www.askthisguy.com/blog/rag-tutorial/schema-rag-agentique.svg Schéma d'un RAG agentique : l'agent orchestrateur interroge documents, bases de données, web et outils métier, évalue s'il a assez d'éléments, relance une recherche ciblée si besoin, puis produit une réponse sourcée https://www.askthisguy.com/fr/blog/open-source-vs-open-weights-ai/ 2026-11-06 https://www.askthisguy.com/blog/open-source-vs-open-weights-ai/face-montage.webp https://www.askthisguy.com/blog/open-source-vs-open-weights-ai/image.webp Illustration generated by Gemini https://www.askthisguy.com/fr/blog/souverainete-numerique-dependance-americaine/ 2025-12-10 https://www.askthisguy.com/blog/souverainete-numerique-dependance-americaine/image.webp https://www.askthisguy.com/blog/souverainete-numerique-dependance-americaine/gemplus.webp Crédit image : Proton https://www.askthisguy.com/en/blog/qwen-4-vs-qwen-3-8-27b/ 2026-11-10 https://www.askthisguy.com/blog/qwen-4-vs-qwen-3-8-27b/amd-qwen-3-8-27b-ryzen-radeon.webp https://www.askthisguy.com/blog/qwen-4-vs-qwen-3-8-27b/rythme-sorties-qwen-2026-en.webp Timeline of Qwen releases in 2026, on four rows. Dense 27B with open weights: Qwen 3.5 27B on February 24, Qwen 3.6 27B on April 22 (two months later), Qwen 3.8 27B on August 14 (nearly four months later). Other open models: Qwen 3.5 397B on February 16, Qwen 3.6 35B-A3B on April 16, Qwen 3.8 2.4T under a custom license on August 12, Qwen 3.8 Flash-Next, a Qwen 4 preview, on August 26. Closed models: Qwen 3.7 Max on May 20, Qwen 3.7 Plus on June 1, Qwen 3.8 Max on August 3. Qwen 4: announced as in training on September 22, "Qwen4 next week" reported on October 29, official teaser on November 6 https://www.askthisguy.com/blog/qwen-4-vs-qwen-3-8-27b/benchmarks-qwen-27b-en.webp Bar chart of scores published by Qwen, out of 100, for Qwen 3.6 27B, Qwen 3.8 27B and Opus 4.6 Max: SWE-bench Pro 53.5, 61.7 and 53.4; Terminal Bench 2.1 63.4, 73.0 and 78.2; OSWorld-Verified 63.9, 84.3 and 72.7; IFBench 69.1, 79.5 and 62.5; GPQA Diamond 87.8, 89.2 and 91.3; HLE 24.0, 30.8 and 40.0 https://www.askthisguy.com/blog/qwen-4-vs-qwen-3-8-27b/poids-vs-intelligence-en.webp Scatter plot of total parameters on a log scale against the Artificial Analysis index v5.0.1. Qwen 3.8 27B, dense, 27 billion parameters, score 20, the only model in the zone that fits on one 80 GB GPU. Mistral Small 4, 119-billion-parameter MoE, score 4. Mistral Medium 3.5, 128-billion-parameter dense model, score 5. Mistral Large 4, MoE of about 1,000 billion parameters, score 26. Qwen 3.8 2.4T, 2,400-billion-parameter MoE, placed at 34, the score of its API version Qwen 3.8 Max. An arrow shows that the top left is better: fewer parameters, higher score https://www.askthisguy.com/en/blog/ia-open-source-licences/ 2026-11-10 https://www.askthisguy.com/blog/ia-open-source-licences/giammarco-boscaro-zeh-ljawhtg-unsplash.webp https://www.askthisguy.com/en/blog/text-to-sql-agent/ 2026-11-09 https://www.askthisguy.com/blog/text-to-sql-agent/multi-database-sql-agent.webp https://www.askthisguy.com/en/blog/reranking-rag/ 2026-11-06 https://www.askthisguy.com/blog/reranking-rag/pierre-bamin-lm4-nmcj4xk-unsplash.webp https://www.askthisguy.com/en/blog/choisis-un-nombre-entre-1-et-10/ 2026-11-05 https://www.askthisguy.com/blog/choisis-un-nombre-entre-1-et-10/riho-kroll-m4sgyahyn5o-unsplash.webp https://www.askthisguy.com/blog/choisis-un-nombre-entre-1-et-10/part-du-7-par-modele-en.webp Share of 7s per model at temperature 1, in French and in English: 100% for Claude Haiku 5.5, GPT-6 Luna, Gemini 3.8 Flash and Gemma 4 31B, 96 to 100% for Qwen 3.8 27B, 74 to 76% for Mistral Large 4 and 65 to 70% for Mistral Small; humans are at 33% and chance at 10% https://www.askthisguy.com/blog/choisis-un-nombre-entre-1-et-10/temperature-part-du-7-en.webp Share of 7s by temperature, prompt in English: Gemini 3.8 Flash, GPT-6 Luna and Gemma 4 31B stay at 100% from 0 to 1.5; Qwen 3.8 27B drops to about 90% at 1.5; Mistral Large 4 falls from 93% to about 72%; Mistral Small falls from 100% to 45%; chance is at 10% https://www.askthisguy.com/en/blog/audit-rag/ 2026-11-04 https://www.askthisguy.com/blog/audit-rag/berke-citak-qv5voi18ld8-unsplash.webp https://www.askthisguy.com/en/blog/forward-deployed-engineer-definition/ 2026-11-03 https://www.askthisguy.com/blog/forward-deployed-engineer-definition/vitaly-gariev-cn54mbf1f8i-unsplash.webp https://www.askthisguy.com/en/blog/mixture-of-experts-dense-vs-moe/ 2026-10-28 https://www.askthisguy.com/blog/mixture-of-experts-dense-vs-moe/fotoflo-iq3c6trhsow-unsplash.webp https://www.askthisguy.com/en/blog/data-room-ia-mcp/ 2026-10-27 https://www.askthisguy.com/blog/data-room-ia-mcp/cover.webp https://www.askthisguy.com/en/blog/ia-due-diligence-trouver-le-bon-document/ 2026-10-23 https://www.askthisguy.com/blog/ia-due-diligence-trouver-le-bon-document/couverture.webp https://www.askthisguy.com/blog/ia-due-diligence-trouver-le-bon-document/trouver-le-document-escalier-en.webp Bar chart: GPT-4 Turbo answers 9% of questions correctly with no document, 19% when it searches the whole filings base, 50% with the right filing in chunks, 79% with the whole right filing, 85% with the right page https://www.askthisguy.com/blog/ia-due-diligence-trouver-le-bon-document/logiciel-vs-modele-en.webp Two-part chart. Left, same model and different software: Mistral Medium 3.5 goes from 22.7% to 70% then 78.7%, GLM-5.2 from 26.7% to 79.3% then 86%. Right, same software and different model at Dewey: 62.9% for GPT-5.4, 83.7% for Claude Opus 4.6 https://www.askthisguy.com/en/blog/chatgpt-enterprise-credits/ 2026-11-06 https://www.askthisguy.com/blog/chatgpt-enterprise-credits/portefeuille-vide.webp https://www.askthisguy.com/blog/chatgpt-enterprise-credits/prix-intelligence-modeles-en.webp Two bar charts aligned by model. Left, the price for the same tokens: GPT-6 Luna 1, GPT-5.6 Luna 2, GPT-6.1 Sol 20, GPT-6 Sol 20, GPT-5.6 Terra 20 to 24, GPT-5.6 Sol 40, GPT-6 Astra 100. Right, the intelligence score: 38, 37, 52, 48, 42, 47 and 53. GPT-6 Luna and GPT-6.1 Sol are highlighted as recommended, GPT-6 Astra as exceptional, the others in grey https://www.askthisguy.com/blog/chatgpt-enterprise-credits/changer-modele-chatgpt-excel-en.webp Three-step diagram of the ChatGPT panel in Excel: click the model name "GPT-5.6 Sol" at the bottom of the panel, choose "GPT-6 Luna" in the model list, then keep reasoning effort on "Medium" or lower https://www.askthisguy.com/en/blog/ovhcloud-strategie-ia/ 2026-10-20 https://www.askthisguy.com/blog/ovhcloud-strategie-ia/octave-klaba-ovhcloud-datacenter.webp https://www.askthisguy.com/blog/ovhcloud-strategie-ia/prix-tokens-europe-en.webp Price per million tokens in euros excluding VAT as of October 20, 2026, input then output. gpt-oss-120b: OVHcloud 0.08 and 0.40, Scaleway 0.15 and 0.60, IONOS 0.15 and 0.65. Llama 3.3 70B: OVHcloud 0.67 and 0.67, Scaleway 0.90 and 0.90, IONOS 0.65 and 0.65. Mistral Small 3.2 24B: OVHcloud 0.09 and 0.28, Scaleway 0.15 and 0.35, IONOS 0.10 and 0.30. Qwen3.5-397B: 0.60 and 3.60 at all three https://www.askthisguy.com/blog/ovhcloud-strategie-ia/prix-ovhcloud-umans-en.webp OVHcloud versus Umans as of October 20, 2026, price per million tokens in euros and Artificial Analysis intelligence index. Entry level: gpt-oss-120b at OVHcloud, €0.08 input, €0.40 output, index 12; DeepSeek V4 Flash at Umans, €0.12, €0.25, index 34. High end: Qwen3.5-397B at OVHcloud, €0.60, €3.60, index 18; GLM 5.3 at Umans, €1.25, €3.92, index 45. Umans prices converted from dollars at the ECB rate of October 19, 2026 https://www.askthisguy.com/en/blog/timestral-bitcoin-prevision/ 2026-10-19 https://www.askthisguy.com/blog/timestral-bitcoin-prevision/beatriz-cattel-c55m3wxaob0-unsplash.webp https://www.askthisguy.com/blog/timestral-bitcoin-prevision/prix-volume-par-annee-en.webp Timestral-209M gain year by year: on 1-hour price, from +0.57% in 2019 to -0.05% in 2026; on next-day volume, always positive, +6.3% in 2026 https://www.askthisguy.com/en/blog/claude-mythos-failles-correctifs/ 2026-10-16 https://www.askthisguy.com/blog/claude-mythos-failles-correctifs/mythos-failles-correctifs.webp https://www.askthisguy.com/en/blog/coreweave-dette-dependance-ia/ 2026-10-16 https://www.askthisguy.com/blog/coreweave-dette-dependance-ia/m-w7weqy01k3s-unsplash.webp https://www.askthisguy.com/en/blog/chatgpt-watermark/ 2026-10-15 https://www.askthisguy.com/blog/chatgpt-watermark/code-ecran-sombre.webp https://www.askthisguy.com/en/blog/french-tech-ownership/ 2026-11-06 https://www.askthisguy.com/blog/french-tech-ownership/fr-to-us.webp https://www.askthisguy.com/en/blog/ai-self-replication/ 2026-10-14 https://www.askthisguy.com/blog/ai-self-replication/ai-worms.webp https://www.askthisguy.com/en/blog/anthropic-scion-git/ 2026-11-06 https://www.askthisguy.com/blog/anthropic-scion-git/de-git-a-scion-migration-moderne.webp https://www.askthisguy.com/en/blog/ai-distillation/ 2026-10-13 https://www.askthisguy.com/blog/ai-distillation/francois-hoppet-kqohccwt7tq-unsplash.webp https://www.askthisguy.com/en/blog/llm-gateway/ 2026-11-06 https://www.askthisguy.com/blog/llm-gateway/denys-nevozhai-7nrsvjvalna-unsplash.webp https://www.askthisguy.com/blog/enterprise-ai-sla/success-rate-per-provider.webp Success rate by inference provider: Scaleway 99.87%, Nebius 99.81%, Cerebras 99.78%, DeepInfra 99.76%, Mistral 99.56% https://www.askthisguy.com/blog/llm-gateway/same-model-throughput-price-by-provider.webp Two bar charts. On the left, the gpt-oss-120b model: Nebius serves it at 40 tokens per second and Groq at 500, both at $0.15 input and $0.60 output per million tokens, Cerebras at 3,000 tokens per second for $0.35 and $0.75. On the right, the DeepSeek V4 Pro model costs $0.57 per million tokens at DeepSeek (China), $1.69 at DeepInfra (United States), $2.11 at Umans and $2.21 at Arkane Cloud (France), $2.27 at Nebius (Netherlands). https://www.askthisguy.com/en/blog/rag-how-many-documents/ 2026-10-07 https://www.askthisguy.com/blog/rag-how-many-documents/schema-rag-filtrer-avant-de-chercher.svg https://www.askthisguy.com/en/blog/mistral-ocr-paddleocr-cost/ 2026-09-30 https://www.askthisguy.com/blog/mistral-ocr-paddleocr-cost/viktor-talashuk-05hlfqu8bfw-unsplash.webp https://www.askthisguy.com/blog/mistral-ocr-paddleocr-cost/ocr-cost-million-pages-en.svg Bar chart comparing the cost of OCR for one million pages: $4,000 with Mistral OCR 4.1, $2,000 with Mistral OCR 3 or OCR 4 in batch, $151 with PaddleOCR-VL on an on-demand RTX PRO 6000 GPU at Verda, $76 on the same GPU at spot pricing https://www.askthisguy.com/en/blog/jev-rag-reranking/ 2026-10-15 https://www.askthisguy.com/blog/jev-rag-reranking/daniel-forsman-ph4zjrwf4x8-unsplash.webp https://www.askthisguy.com/en/blog/how-ai-works-key-concepts/ 2026-09-07 https://www.askthisguy.com/blog/how-ai-works-key-concepts/markus-spiske-vic0envgdtu-unsplash.webp https://www.askthisguy.com/blog/how-ai-works-key-concepts/schema-reseau-de-neurones.webp Diagram of a neural network: an input layer receives photos of cats and dogs, the signals travel through several internal layers of interconnected neurons, and the output layer sorts the images into two categories, cats and dogs https://www.askthisguy.com/blog/how-ai-works-key-concepts/intelligence-vs-parametres-modeles.webp Scatter plot comparing the capability level of open models against their parameter count on a logarithmic scale. The curve rises sharply up to around thirty billion parameters, then flattens: beyond that, multiplying size by a hundred gains only a few points https://www.askthisguy.com/blog/how-ai-works-key-concepts/intelligence-index-modeles-par-pays.webp Ranking of models by a general capability index, with the flag of each country of origin. The top two are American, the next two Chinese and very close behind, the French models come in well below, and the Swiss model brings up the rear https://www.askthisguy.com/blog/how-ai-works-key-concepts/tokens-francais-anglais-swahili.webp The same text broken into tokens in three languages. In French, 2,658 characters produce 692 tokens. In English, 2,589 characters produce only 606. In Swahili, 2,688 characters produce 835 https://www.askthisguy.com/blog/how-ai-works-key-concepts/vitesse-fournisseurs-inference.webp Generation speed for the same open model depending on which provider runs it, with each provider's country flag. The top four are American, the fastest exceeding 1,900 tokens per second; the two European providers bring up the rear, at 243 and 159 tokens per second https://www.askthisguy.com/blog/how-ai-works-key-concepts/schema-appel-outil.webp Diagram of a tool call: the user's question reaches the model, which either answers directly or triggers a tool run by the company's own software (ERP, CRM, document base, web search, calculations, sending emails) before formulating its answer https://www.askthisguy.com/en/blog/hybrid-search-retrieval-methods/ 2026-08-26 https://www.askthisguy.com/blog/hybrid-search-retrieval-methods/schema-methodes-recherche-complementaires.svg https://www.askthisguy.com/en/blog/ia-francaise-souverainete/ 2026-10-15 https://www.askthisguy.com/blog/ia-francaise-souverainete/chris-karidis-nnzkZNYWHaU-unsplash.webp https://www.askthisguy.com/en/blog/enterprise-ai-inference/ 2026-10-12 https://www.askthisguy.com/blog/enterprise-ai-inference/ilias-gainutdinov-UILFtxyE9Eg-unsplash.webp https://www.askthisguy.com/blog/enterprise-ai-inference/cout-vs-intelligence-modeles-ia.webp Scatter plot comparing cost per million tokens and intelligence index for four offers. DeepSeek V4 Flash served by the French provider Umans reaches an index of 50 for €0.11 per MTok, while OpenAI's GPT-5.6 Sol reaches 56 for €7.35, and GPT-5.6 Luna 46 for €0.29. https://www.askthisguy.com/en/blog/enterprise-ai-agents/ 2026-08-08 https://www.askthisguy.com/blog/enterprise-ai-agents/campaign-creators-kq4tbklji-unsplash.webp https://www.askthisguy.com/en/blog/internal-company-chatbot/ 2026-08-07 https://www.askthisguy.com/blog/internal-company-chatbot/philipp-katzenberger-iijruoerocq-unsplash.webp https://www.askthisguy.com/blog/internal-company-chatbot/internal-company-chatbot-sources.webp Diagram of an internal company chatbot: sources (Drive, SharePoint, Confluence, Notion, intranet) feed a RAG engine that applies each person's permissions and answers by citing the 2026 leave agreement https://www.askthisguy.com/en/blog/knowledge-management-why-it-fails/ 2026-08-26 https://www.askthisguy.com/blog/knowledge-management-why-it-fails/salle-etude-vide.webp https://www.askthisguy.com/en/blog/enterprise-ai-sla/ 2026-10-12 https://www.askthisguy.com/blog/enterprise-ai-sla/bermix-studio-wbc8mlfuzu-unsplash.webp https://www.askthisguy.com/blog/enterprise-ai-sla/microsoft-availability.webp Microsoft 365 worldwide availability by quarter, around 99.9% and above (source Microsoft Learn) https://www.askthisguy.com/blog/enterprise-ai-sla/success-rate-per-provider.webp Success rate by inference provider: Scaleway 99.87%, Nebius 99.81%, Cerebras 99.78%, DeepInfra 99.76%, Mistral 99.56% https://www.askthisguy.com/en/blog/chatbot-produit-saas/ 2026-11-10 https://www.askthisguy.com/blog/chatbot-produit-saas/ant-rozetsky-r4iD__fTqIs-unsplash.webp https://www.askthisguy.com/blog/chatbot-produit-saas/atg-console-help-sharepoint-en.webp Help widget inside the Ask This Guy admin console, guiding an administrator step by step through configuring the SharePoint connector https://www.askthisguy.com/doc/tags-alerts-01-en.webp Tags & Alerts screen with three example rules: interest in a premium feature, purchase intent, and a customer wanting to cancel https://www.askthisguy.com/en/blog/data-analysis-tools/ 2026-07-08 https://www.askthisguy.com/blog/data-analysis-tools/ux-indonesia-8mikj83lmsq-unsplash.webp https://www.askthisguy.com/blog/data-analysis-tools/illustration-excel.webp Excel spreadsheet with formulas, data and charts https://www.askthisguy.com/blog/data-analysis-tools/illustration-powerbi.webp Power BI dashboard with charts and tracking indicators https://www.askthisguy.com/blog/data-analysis-tools/illustration-jupyter.webp Jupyter notebook with Python code and data visualization https://www.askthisguy.com/blog/data-analysis-tools/illustration-matlab.webp MATLAB interface for numerical modelling and statistical analysis https://www.askthisguy.com/blog/data-analysis-tools/illustration-ask-this-guy.webp Ask This Guy's Talk to Data interface for querying data in natural language https://www.askthisguy.com/en/blog/financial-analysis-without-excel/ 2026-10-15 https://www.askthisguy.com/blog/financial-analysis-without-excel/carlos-muza-hpjSkU2UYSU-unsplash.webp https://www.askthisguy.com/en/blog/qualified-leads-website-chatbot/ 2026-05-29 https://www.askthisguy.com/blog/qualified-leads-website-chatbot/juan-rodriguez-duarte-31NZ7nGupm8-unsplash.webp https://www.askthisguy.com/en/blog/contextual-ai-rag-use-cases/ 2026-05-05 https://www.askthisguy.com/blog/contextual-ai-rag-use-cases/vitaly-gariev-Vvn0OD0IxBM-unsplash.webp https://www.askthisguy.com/en/blog/enterprise-rag-production-mistakes/ 2026-04-24 https://www.askthisguy.com/blog/enterprise-rag-production-mistakes/maksym-kaharlytskyi-Q9y3LRuuxmg-unsplash.webp https://www.askthisguy.com/blog/enterprise-rag-production-mistakes/success-rate-q1-text-generation.webp ATG Q1 2026 benchmark: text generation success rates by provider https://www.askthisguy.com/en/blog/use-minimal-ai-automation/ 2026-03-26 https://www.askthisguy.com/blog/use-minimal-ai-automation/paul-lichtblau--n_bYa2Bhwg-unsplash.webp https://www.askthisguy.com/en/blog/databases-gold-mine-ai-shovels/ 2026-03-18 https://www.askthisguy.com/blog/databases-gold-mine-ai-shovels/gold-mine.webp https://www.askthisguy.com/blog/databases-gold-mine-ai-shovels/you-dig.webp Illustration: “In this world there are two kinds of people: those with loaded guns and those who dig. You dig.” https://www.askthisguy.com/blog/databases-gold-mine-ai-shovels/claudio-schwarz-fyeOxvYvIyY-unsplash.webp Credit: Claudio Schwarz https://www.askthisguy.com/en/blog/llm-local-enterprise-sovereignty-illusion/ 2026-10-15 https://www.askthisguy.com/blog/llm-local-enterprise-sovereignty-illusion/data-center.webp https://www.askthisguy.com/blog/llm-local-enterprise-sovereignty-illusion/agentic-ai-dobi-2.webp Agentic AI: chaining steps and tool calls https://www.askthisguy.com/blog/llm-local-enterprise-sovereignty-illusion/small-is-beautiful-2.webp When small models are enough https://www.askthisguy.com/en/blog/how-mcp-accelerate-agentic-ai/ 2026-10-15 https://www.askthisguy.com/blog/how-mcp-accelerate-agentic-ai/water-connection.webp https://www.askthisguy.com/blog/how-mcp-accelerate-agentic-ai/mermaid-diagram-tool-calling-en.webp AI agent tool calling flow https://www.askthisguy.com/blog/how-mcp-accelerate-agentic-ai/tool-discovery-selection-en.webp MCP tool discovery and selection process https://www.askthisguy.com/blog/how-mcp-accelerate-agentic-ai/mcp-ui-flow.webp MCP tool configuration and selection in the admin interface https://www.askthisguy.com/blog/how-mcp-accelerate-agentic-ai/user-surf-chat.webp Ask This Guy Chat UI https://www.askthisguy.com/en/blog/embedding-strategy/ 2026-01-28 https://www.askthisguy.com/blog/embedding-strategy/chuttersnap-cGXdjyP6-NU-unsplash.webp https://www.askthisguy.com/blog/embedding-strategy/mistral_openai_table.webp Mistral & OpenAI performance https://www.askthisguy.com/blog/embedding-strategy/mistral_openai_1s.webp Mistral & OpenAI performance https://www.askthisguy.com/blog/embedding-strategy/providers_comparison.webp Embedding providers performance comparison https://www.askthisguy.com/blog/embedding-strategy/european_providers_1s.webp Performance comparison of embedding providers https://www.askthisguy.com/en/blog/rag-tutorial/ 2026-10-12 https://www.askthisguy.com/blog/rag-tutorial/scott-graham-5fNmWej4tAA-unsplash%20(1).webp https://www.askthisguy.com/blog/rag-tutorial/ai-lost.webp AI unable to answer questions about internal company documents https://www.askthisguy.com/blog/rag-tutorial/bibliothecaire.webp A fast and precise library https://www.askthisguy.com/blog/rag-tutorial/workflow-en.svg Simplified RAG workflow https://www.askthisguy.com/en/blog/open-source-vs-open-weights-ai/ 2026-11-06 https://www.askthisguy.com/blog/open-source-vs-open-weights-ai/face-montage.webp https://www.askthisguy.com/blog/open-source-vs-open-weights-ai/image.webp Illustration generated by Gemini https://www.askthisguy.com/en/blog/souverainete-numerique-dependance-americaine/ 2025-12-10 https://www.askthisguy.com/blog/souverainete-numerique-dependance-americaine/image.webp https://www.askthisguy.com/fr/blog/tags/ 2026-11-10 https://www.askthisguy.com/fr/blog/tag/gestion-connaissances/ 2026-11-06 https://www.askthisguy.com/fr/blog/tag/embeddings/ 2026-11-06 https://www.askthisguy.com/fr/blog/tag/llm/ 2026-11-10 https://www.askthisguy.com/fr/blog/tag/open-weights/ 2026-11-10 https://www.askthisguy.com/fr/blog/tag/agents-ia/ 2026-11-10 https://www.askthisguy.com/fr/blog/tag/mcp/ 2026-11-09 https://www.askthisguy.com/fr/blog/tag/infrastructure-ia/ 2026-11-10 https://www.askthisguy.com/fr/blog/tag/souverainete/ 2026-11-10 https://www.askthisguy.com/fr/blog/tag/rgpd/ 2026-11-06 https://www.askthisguy.com/fr/blog/tag/securite-ia/ 2026-10-16 https://www.askthisguy.com/fr/blog/tag/roi/ 2026-11-10 https://www.askthisguy.com/fr/blog/tag/analyse-de-donnees/ 2026-11-10 https://www.askthisguy.com/fr/blog/tag/pme/ 2026-09-07 https://www.askthisguy.com/fr/blog/tag/mistral/ 2026-11-10 https://www.askthisguy.com/fr/blog/tag/openai/ 2026-11-06 https://www.askthisguy.com/fr/blog/tag/anthropic/ 2026-11-06 https://www.askthisguy.com/en/blog/tags/ 2026-11-10 https://www.askthisguy.com/en/blog/tag/gestion-connaissances/ 2026-11-06 https://www.askthisguy.com/en/blog/tag/embeddings/ 2026-11-06 https://www.askthisguy.com/en/blog/tag/llm/ 2026-11-10 https://www.askthisguy.com/en/blog/tag/open-weights/ 2026-11-10 https://www.askthisguy.com/en/blog/tag/agents-ia/ 2026-11-10 https://www.askthisguy.com/en/blog/tag/mcp/ 2026-11-09 https://www.askthisguy.com/en/blog/tag/infrastructure-ia/ 2026-11-10 https://www.askthisguy.com/en/blog/tag/souverainete/ 2026-11-10 https://www.askthisguy.com/en/blog/tag/rgpd/ 2026-11-06 https://www.askthisguy.com/en/blog/tag/securite-ia/ 2026-10-16 https://www.askthisguy.com/en/blog/tag/roi/ 2026-11-10 https://www.askthisguy.com/en/blog/tag/analyse-de-donnees/ 2026-11-10 https://www.askthisguy.com/en/blog/tag/pme/ 2026-08-26 https://www.askthisguy.com/en/blog/tag/mistral/ 2026-11-10 https://www.askthisguy.com/en/blog/tag/openai/ 2026-11-06 https://www.askthisguy.com/en/blog/tag/anthropic/ 2026-11-06 https://www.askthisguy.com/fr/solutions/ 2026-11-10 https://www.askthisguy.com/fr/solutions/agentic-automation/ 2026-11-10 https://www.askthisguy.com/fr/solutions/ai-consulting/ 2026-11-04 https://www.askthisguy.com/fr/solutions/ai-custom/ 2026-11-10 https://www.askthisguy.com/fr/solutions/finance/ 2026-11-10 https://www.askthisguy.com/fr/solutions/forward-deployed-engineer/ 2026-11-06 https://www.askthisguy.com/fr/solutions/knowledge-rag/ 2026-11-10 https://www.askthisguy.com/fr/solutions/partners/ 2026-11-10 https://www.askthisguy.com/fr/solutions/product-assistant/ 2026-11-10 https://www.askthisguy.com/fr/solutions/talk-to-data/ 2026-11-10 https://www.askthisguy.com/fr/solutions/website-chatbot/ 2026-10-28 https://www.askthisguy.com/fr/solutions/formation-ia/ 2026-11-10 https://www.askthisguy.com/fr/solutions/prototypage-ia/ 2026-11-10 https://www.askthisguy.com/en/solutions/ 2026-11-10 https://www.askthisguy.com/en/solutions/agentic-automation/ 2026-11-10 https://www.askthisguy.com/en/solutions/ai-consulting/ 2026-11-04 https://www.askthisguy.com/en/solutions/ai-custom/ 2026-11-10 https://www.askthisguy.com/en/solutions/finance/ 2026-11-10 https://www.askthisguy.com/en/solutions/forward-deployed-engineer/ 2026-11-06 https://www.askthisguy.com/en/solutions/knowledge-rag/ 2026-11-10 https://www.askthisguy.com/en/solutions/partners/ 2026-11-10 https://www.askthisguy.com/en/solutions/product-assistant/ 2026-11-10 https://www.askthisguy.com/en/solutions/talk-to-data/ 2026-11-10 https://www.askthisguy.com/en/solutions/website-chatbot/ 2026-10-28 https://www.askthisguy.com/en/solutions/ai-training/ 2026-11-10 https://www.askthisguy.com/en/solutions/ai-proof-of-concept/ 2026-11-10 https://www.askthisguy.com/fr/solutions/secteurs/ 2026-11-02 https://www.askthisguy.com/fr/solutions/secteurs/finance/ 2026-11-10 https://www.askthisguy.com/fr/solutions/secteurs/editeurs-logiciels/ 2026-11-10 https://www.askthisguy.com/en/solutions/industries/ 2026-11-02 https://www.askthisguy.com/en/solutions/industries/finance/ 2026-11-10 https://www.askthisguy.com/en/solutions/industries/software-vendors/ 2026-11-10 https://www.askthisguy.com/fr/cas-clients/ 2026-11-10 https://www.askthisguy.com/fr/cas-clients/ucopia-ia-developpement/ 2026-11-10 https://www.askthisguy.com/fr/cas-clients/chatbot-portail-partenaires/ 2026-11-10 https://www.askthisguy.com/en/case-studies/ 2026-11-10 https://www.askthisguy.com/en/case-studies/ucopia-ai-development/ 2026-11-10 https://www.askthisguy.com/en/case-studies/partner-portal-chatbot/ 2026-11-10 https://www.askthisguy.com/fr/benchmarks/ 2026-11-10 https://www.askthisguy.com/fr/benchmarks/financebench/ 2026-11-10 https://www.askthisguy.com/en/benchmarks/ 2026-11-10 https://www.askthisguy.com/en/benchmarks/financebench/ 2026-11-10 https://www.askthisguy.com/fr/videos/ 2026-11-10 https://www.askthisguy.com/en/videos/ 2026-11-10 https://www.askthisguy.com/privacy.html 2026-10-06 https://www.askthisguy.com/privacy-en.html 2026-10-06 https://www.askthisguy.com/terms.html 2026-08-08 https://www.askthisguy.com/terms-en.html 2026-08-08