Journal des modifications des modèles d'IA · 2026-09-15

85 événements enregistrés le 2026-09-15.

Modèle ajouté (25)

nano-gpt New model: Gemini 2.5 Pro Preview 0506
nano-gpt New model: Claude Haiku 4.5 Thinking
nano-gpt New model: Claude 4.5 Opus Thinking
nano-gpt New model: Claude Sonnet 4.5 Thinking
nano-gpt New model: Schematron V2 Small
nano-gpt New model: Schematron V2 Turbo
llmgateway-providers New model: Atria Dawn Preview (Atria)
llmgateway-providers New model: DeepSeek V4.1 Flash (Alibaba Cloud)
infer New model: GPT-5.6 Sol (Official API)
infer New model: GPT-6 Astra (Official API)
aki-io New model: GLM-5.3
vercel New model: Fugu Max
vercel New model: Fugu Ultra v2
vercel New model: Ling 3.0 Flash VL (Free)
vercel New model: Ling 3.0 Flash VL
llmgateway New model: Atria Dawn Preview
kilo New model: DeepSeek: DeepSeek Pro Latest
kilo New model: DeepSeek: DeepSeek Flash Latest
openrouter New model: DeepSeek Pro Latest
openrouter New model: DeepSeek Flash Latest
amd New model: MiniCPM5-2B
vispark New model: Vision Large
vispark New model: Vision Medium
vispark New model: Vision Small
edenai New model: DeepSeek V4.1 Flash (Alibaba)

Prix modifié (24)

Alibaba (Qwen)Price update: Qwen 3.6 Plus Uncensored now $0.5/$3 (was $0.325/$1.95)
Alibaba (Qwen)Price update: Qwen3.8 Flash Next now $0.15/$0.47 (was $0.12/$0.4)
Mistral Price update: Mistral Medium 3.5 now $1.5/$7.5 (was $1.393/$7.13)
cortecs Price update: ministral-14b-2512 now $0.24/$0.24 (was $0.223/$0.223)
cortecs Price update: ministral-3b-2512 now $0.123/$0.123 (was $0.111/$0.111)
cortecs Price update: ministral-8b-2512 now $0.179/$0.179 (was $0.167/$0.167)
cortecs Price update: voxtral-small-2507 now $0.123/$0.368 (was $0.111/$0.334)
DeepSeek Price update: DeepSeek V4.1 Flash now $0.14/$0.28 (was $0.15/$0.6)
DeepSeek Price update: DeepSeek V4 Flash Latest now $0.04/$0.1 (was $0.0352/$0.1056)
edenai Price update: GPT OSS 120B (FlexAI) now $0.173265/$0.69306 (was $0.17388/$0.69552)
edenai Price update: Ministral 3 14B (Infomaniak) now $0.34653/$0.46204 (was $0.34776/$0.46368)
edenai Price update: DeepSeek V4 Flash 0731 (Alibaba) now $0.22/$0.66 (was $0.176/$0.528)
edenai Price update: DeepSeek V4 Pro 0813 (Alibaba) now $0.66/$1.98 (was $0.5808/$1.7424)
Google Price update: Gemma 4 31B now $0.14/$0.4 (was $0.4/$1)
hyper Price update: GPT OSS 120B now $0.178/$0.68 (was $0.168/$0.66)
Mistral Price update: Mistral Small 3.1 24B now $0.1/$0.3 (was $0.351/$0.555)
nano-gpt Price update: GLM 5.3 Flash Uncensored now $0.2/$0.8 (was $0.35/$1.4)
nano-gpt Price update: DeepSeek V4.1 Flash Thinking now $0.1/$0.4 (was $0.15/$0.6)
ollama-cloud Price update: mistral-large-3:675b now $0.5/$1.5 (was $?/$?)
ollama-cloud Price update: nemotron-3-nano:30b now $0.06/$0.24 (was $?/$?)
ollama-cloud Price update: nemotron-3-super now $0.015/$0.6 (was $?/$?)
ollama-cloud Price update: qwen3.5:397b now $0.6/$3.6 (was $?/$?)
vercel Price update: Ling 3.0 Flash now $0.021/$0.063 (was $0.06/$0.18)
Z.AI / Zhipu Price update: GLM Latest now $0.9/$3 (was $0.936/$3.168)

Contexte modifié (6)

Alibaba (Qwen)Context window: Qwen 3.6 Plus Uncensored now 1,000,000 (was 991,808)
Alibaba (Qwen)Context window: Qwen3.8 Flash Next now 262,144 (was 1,048,576)
ByteDance (Doubao)Context window: Seed 2.1 Turbo now 262,144 (was 256,000)
Mistral Context window: Devstral 2 now 200,000 (was 262,144)
Moonshot AI Context window: kimi-for-coding now 1,048,576 (was 262,144)
openrouter Context window: Nemotron 3 Super (free) now 262,144 (was 1,000,000)

Capacité mise à jour (21)

Alibaba (Qwen)Capabilities updated: Qwen 3.6 Plus Uncensored
Alibaba (Qwen)Capabilities updated: Qwen3.8 Flash Next
amazon-bedrock Capabilities updated: GPT-5.6 Luna (Global)
amazon-bedrock Capabilities updated: GPT-5.6 Sol (Global)
amazon-bedrock Capabilities updated: GPT-5.6 Terra (Global)
amazon-bedrock Capabilities updated: GPT-6 Astra (Global)
amazon-bedrock Capabilities updated: GPT-5.6 Luna (US)
amazon-bedrock Capabilities updated: GPT-5.6 Sol (US)
amazon-bedrock Capabilities updated: GPT-5.6 Terra (US)
amazon-bedrock Capabilities updated: GPT-6 Astra (US)
ByteDance (Doubao)Capabilities updated: Seed 2.1 Turbo
kilo Capabilities updated: Qwen: Qwen3.8 27B
Mistral Capabilities updated: Mistral: Mistral 7B Instruct v0.1
Mistral Capabilities updated: Devstral 2
Moonshot AI Capabilities updated: kimi-for-coding
nano-gpt Capabilities updated: The Drummer Cydonia 24B v4.1
openrouter Capabilities updated: Ling-3.0-flash (free)
openrouter Capabilities updated: Nex-N2-Pro (free)
openrouter Capabilities updated: Qwen3.8 27B
xAI Capabilities updated: Grok 3 Beta
xAI Capabilities updated: Grok 3 Mini Beta

Limite de sortie modifiée (9)

Alibaba (Qwen)Max output: Qwen3.8 Flash Next updated
ByteDance (Doubao)Max output: Seed 2.1 Turbo updated
DeepSeek Max output: DeepSeek V4 Flash Latest updated
Meta Max output: Llama 4 Maverick updated
Mistral Max output: Mistral: Mistral 7B Instruct v0.1 updated
Mistral Max output: Devstral 2 updated
nano-gpt Max output: Granite 4.1 8B updated
openrouter Max output: Nex-N2-Pro (free) updated
Z.AI / Zhipu Max output: GLM Latest updated

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