Journal des modifications des modèles d'IA · 2026-09-07

37 événements enregistrés le 2026-09-07.

Modèle ajouté (8)

Alibaba (Qwen)New model: Qwen3.8 27B
nano-gpt New model: TheDrummer/Artemis v1.1
neuralwatt New model: Gemma 4 31B
abacus New model: Kimi K2 Turbo Preview
302ai New model: kimi-k2-0905-preview
302ai New model: kimi-k2-thinking-turbo
zenmux New model: Kimi K2 Thinking Turbo
edenai New model: Inkling (Databricks)

Limite de sortie modifiée (8)

DeepSeek Max output: DeepSeek V4 Flash 0731 updated
DeepSeek Max output: DeepSeek V4 Flash Latest updated
Mistral Max output: Mistral: Mistral 7B Instruct v0.1 updated
Mistral Max output: Devstral 2 updated
Moonshot AI Max output: Kimi K2 Thinking updated
nano-gpt Max output: GPT-5.1 (2025-11-13) updated
nano-gpt Max output: Granite 4.1 8B updated
openrouter Max output: Nex-N2-Pro (free) updated

Prix modifié (7)

DeepSeek Price update: DeepSeek V4 Flash Latest now $0.045/$0.09 (was $0.04998/$0.09996)
Moonshot AI Price update: Kimi K2.5 now $0.3/$1.9 (was $0.6/$3)
Moonshot AI Price update: Kimi K2 Thinking now $0.47/$2 (was $0.6/$2.5)
venice Price update: GPT-5.6 Sol now $2.5/$12.5 (was $6.25/$37.5)
venice Price update: GPT-5.6 Sol Pro now $2.5/$12.5 (was $6.25/$37.5)
venice Price update: GPT-5.6 Terra now $2.5/$15 (was $3.125/$18.75)
venice Price update: GPT-5.6 Terra Pro now $2.5/$15 (was $3.125/$18.75)

Contexte modifié (4)

Mistral Context window: Devstral 2 now 200,000 (was 262,144)
Moonshot AI Context window: Kimi K2 Thinking now 262,143 (was 262,144)
nano-gpt Context window: GPT-5.1 (2025-11-13) now 400,000 (was 1,000,000)
openrouter Context window: Nemotron 3 Super (free) now 262,144 (was 1,000,000)

Pricing and capabilities are refreshed daily and reconciled against each provider's official documentation. Always verify critical production decisions with the provider directly.

Modèle ajouté (8)

Limite de sortie modifiée (8)

Prix modifié (7)

Capacité mise à jour (10)

Contexte modifié (4)