Journal des modifications des modèles d'IA · 2026-08-29

82 événements enregistrés le 2026-08-29.

Modèle ajouté (34)

ByteDance (Doubao)New model: Seed 2.1 Turbo
ByteDance (Doubao)New model: Seed Evolving
cortecs New model: Qwen3.8 Flash Next
nano-gpt New model: Abliterated Model Large
nano-gpt New model: Abliterated Model
nano-gpt New model: Tencent Hy4 Preview
nano-gpt New model: Moonlight Dusk
nano-gpt New model: Luminous Mirror
nano-gpt New model: Isometry
nano-gpt New model: Musica
nano-gpt New model: Opus Distill
nano-gpt New model: DarkIdol
nano-gpt New model: Shadow Siren
nano-gpt New model: Novelist
nano-gpt New model: Chimera X
nano-gpt New model: Fabled
nano-gpt New model: Garnet
nano-gpt New model: Dark Soul
nano-gpt New model: Gembrain
nano-gpt New model: Qwen 3.8 27B Fable
tencent-tokenhub New model: Hy4 preview
openrouter New model: Hy4 preview
opencode-go New model: Hy4 preview
novita New model: Qwen3.8 Flash (NovitaAI)
novita New model: GLM-5.3 (NovitaAI)
novita New model: Qwen3.8 27B (NovitaAI)
opencode New model: Ling 3.0 Flash Fin Free
requesty New model: GLM-5.3 (EU)
amd New model: Qwen3.8 Flash Next
vercel New model: Tencent Hy4 Preview
vercel New model: Wan v3.0 Video Prime
tencent-token-plan New model: Hy4 preview
kilo New model: Hy4 preview
fireworks-ai New model: GLM 5.3

Limite de sortie modifiée (8)

Alibaba (Qwen)Max output: Qwen3 VL 30B A3B Instruct updated
Meta Max output: Muse Glimmer 30B updated
Mistral Max output: Mistral: Mistral 7B Instruct v0.1 updated
Mistral Max output: Devstral 2 updated
NVIDIA Max output: Nemotron 3.5 Lightning 30B A3B updated
openrouter Max output: Nex-N2-Pro (free) updated
vercel Max output: Hy3 updated
Z.AI / Zhipu Max output: GLM Latest updated

Prix modifié (13)

DeepSeek Price update: DeepSeek V4 Flash 0731 now $0.045/$0.09 (was $0.06/$0.12)
DeepSeek Price update: DeepSeek V4 Pro 0813 now $0.35/$0.8 (was $0.442/$0.884)
nano-gpt Price update: Claw High now $1/$3.2 (was $1.4/$4.4)
nano-gpt Price update: Claw Low now $1/$3.2 (was $1.4/$4.4)
nano-gpt Price update: Claw Medium now $1/$3.2 (was $1.4/$4.4)
nano-gpt Price update: GLM 5.3 Thinking now $1/$3.2 (was $1.4/$4.4)
nano-gpt Price update: Hermes High now $1/$3.2 (was $1.4/$4.4)
nano-gpt Price update: Hermes Low now $1/$3.2 (was $1.4/$4.4)
nano-gpt Price update: Hermes Medium now $1/$3.2 (was $1.4/$4.4)
NVIDIA Price update: Nemotron 3.5 Lightning 30B A3B now $0.05/$0.2 (was $0.05/$0.15)
OpenAI Price update: GPT-5.6 Luna Pro now $0.2/$1.2 (was $0.1/$0.6)
OpenAI Price update: GPT-5.6 Terra Pro now $2/$12 (was $1/$6)
vercel Price update: Hy3 now $0.14/$0.58 (was $0.132/$0.528)

Capacité mise à jour (23)

kilo Capabilities updated: Qwen: Qwen3.8 27B
Mistral Capabilities updated: Mistral: Mistral 7B Instruct v0.1
Mistral Capabilities updated: Devstral 2
ofox Capabilities updated: Seed 1.6 Flash
ofox Capabilities updated: Seed 1.6 Vision
ofox Capabilities updated: Seed 2.1 Pro
ofox Capabilities updated: Seed Character
openrouter Capabilities updated: Ling-3.0-flash (free)
openrouter Capabilities updated: Nex-N2-Pro (free)
openrouter Capabilities updated: Qwen3.8 27B
volcengine Capabilities updated: Seed 1.6
volcengine Capabilities updated: Seed 1.6 Flash
volcengine Capabilities updated: Seed 1.6 Vision
volcengine Capabilities updated: Seed 1.8
volcengine Capabilities updated: Seed 2.0 Lite
volcengine Capabilities updated: Seed 2.0 Mini
volcengine Capabilities updated: Seed 2.0 Pro
volcengine Capabilities updated: Seed 2.1 Pro
volcengine Capabilities updated: Seed 2.1 Turbo
volcengine Capabilities updated: Seed Character
xAI Capabilities updated: Grok 3 Beta
xAI Capabilities updated: Grok 3 Mini Beta
Z.AI / Zhipu Capabilities updated: GLM Latest

Contexte modifié (4)

Mistral Context window: Devstral 2 now 200,000 (was 262,144)
openrouter Context window: Nemotron 3 Super (free) now 262,144 (was 1,000,000)
vercel Context window: Hy3 now 262,144 (was 256,000)
Z.AI / Zhipu Context window: GLM Latest now 1,310,720 (was 1,048,576)

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