News-Tracker

Modeller

AI & LLM · daglig briefing

Ændringer siden forrige snapshot

Ny i indekset

  • Mistral Large 4 Preview er ny i Artificial Analysis' indeks (intelligence 25.3, Mistral).
  • Mistral Small 4 (Reasoning) er ny i Artificial Analysis' indeks (intelligence 4.3, Mistral).

Score ændret

  • Mistral Large 4 (High): intelligence 25.3 → 26 (+0.7).

Ny i HF-trending

  • mistralai/Mistral-Large-4-1T-A52B er ny i Hugging Faces trending (score 311).
  • TenStrip/10Eros-Max er ny i Hugging Faces trending (score 74).
  • openbmb/MiniCPM-OCR-Jev er ny i Hugging Faces trending (score 71).
  • facebook/sam-3d-body-dinov3 er ny i Hugging Faces trending (score 64).
  • audnai/penclaw-Kimi-K3.1-pro-unrestricted er ny i Hugging Faces trending (score 62).

Intelligence Index — Artificial Analysis

#ModelIntel.Agentic CreatorUdgivet
1 Gemini 4 Carbon (High) 59.4 Google 2026-10-28
2 Claude Opus 5.5 (Max, Default Fallback) 5 effort-varianter 58.4 Anthropic 2026-09-22
3 GPT-6 Astra (Max) 5 effort-varianter 56.9 OpenAI 2026-09-03
4 GPT-6.1 Sol (Max) 5 effort-varianter 54.5 OpenAI 2026-09-29
5 Claude Sonnet 5.5 (Max, Default Fallback) 5 effort-varianter 54.2 Anthropic 2026-09-28
6 GPT-6.1 Terra (Max) 6 effort-varianter 51 OpenAI 2026-10-26
7 Gemini 4 Argon (High) 50.4 Google 2026-09-30
8 Claude Fable 5.1 (Xhigh, Default Fallback) 5 effort-varianter 49.7 Anthropic 2026-09-01
9 Muse Wonder Full (Max) 49.7 Meta 2026-10-12
10 Kimi K3.1 Pro ⬇ hf.co/moonshotai/Kimi-K3.1-Pro 46 Kimi 2026-10-23
11 Muse Wonder (Max) ⬇ hf.co/meta-01/muse-wonder 43.3 Meta 2026-10-12
12 Muse Spark 1.3 (Max) 2 effort-varianter 36.4 Meta 2026-09-02
13 Claude Haiku 5.5 (Max) 5 effort-varianter 36.2 Anthropic 2026-10-07
14 Grok 4.7 (High) 3 effort-varianter 35 SpaceXAI 2026-09-21
15 Qwen3.8 Max (0902) 2 effort-varianter 33.6 Alibaba 2026-09-02
16 DeepSeek V4.1 Flash 1017 (Max) ⬇ hf.co/deepseek-ai/DeepSeek-V4.1-Flash-1017 32.5 DeepSeek 2026-10-17
17 GLM-5.3 (Max) 2 effort-varianter ⬇ hf.co/zai-org/GLM-5.3 31.7 Z AI 2026-08-18
18 Gemini 3.8 Flash (High) 3 effort-varianter 30.5 Google 2026-09-02
19 MiMo-V2.6-Pro 29.9 Xiaomi 2026-09-21
20 Ling 3.1 Flash 1019 ⬇ hf.co/inclusionAI/Ling-3.1-flash-1019 29.5 InclusionAI 2026-10-22
21 MiniMax-M3.1-Flash ⬇ hf.co/MiniMaxAI/MiniMax-M3.1-Flash 28.9 MiniMax 2026-10-19
22 Step 5 Preview ⬇ hf.co/stepfun-ai/Step-5-Preview 28.5 StepFun 2026-09-18
23 GLM-5.3-Flash ⬇ hf.co/zai-org/GLM-5.3-Flash 27.8 Z AI 2026-08-26
24 GPT-6 Luna (Max) 6 effort-varianter 27.8 OpenAI 2026-09-22
25 MiMo-V2.6-Flash 26.7 Xiaomi 2026-09-21
26 Mistral Large 4 (High) 2 effort-varianter 26 Mistral 2026-10-29
27 Mistral Large 4 Preview 25.3 Mistral 2026-10-06
28 Muse Spark Open (Xhigh) 24.7 Meta 2026-10-15
29 DeepSeek V4 Pro 0813 (Max) 2 effort-varianter ⬇ hf.co/deepseek-ai/DeepSeek-V4-Pro 23.1 DeepSeek 2026-08-13
30 Qwen3.8 27B (Xhigh) 4 effort-varianter ⬇ hf.co/Qwen/Qwen3.8-27B 20.1 Alibaba 2026-08-14
31 MiniMax-M3 ⬇ hf.co/MiniMaxAI/MiniMax-M3 18.7 MiniMax 2026-06-01
32 K2 Horizon 375B A23B ⬇ hf.co/IFM/K2-Horizon-375B-A23B 17.8 Institute of Foundation Models 2026-09-03
33 Beam 1.0 17.7 Reflection 2026-10-16
34 Kimi K3 (Low) 2 effort-varianter 17.4 Kimi 2026-07-16
35 Nemotron 3.5 Super VL 13.6 NVIDIA 2026-10-15
36 Inkling Small ⬇ hf.co/thinkingmachines/Inkling-Small 12.9 Thinking Machines 2026-07-30
37 Inkling (Xhigh) ⬇ hf.co/thinkingmachines/Inkling 12.5 Thinking Machines 2026-07-15
38 Nemotron 3 Ultra 550B A55B (Reasoning) 11.8 NVIDIA 2026-06-04
39 Gemini 3.5 Flash-Lite 11.7 Google 2026-07-21
40 Muse Glimmer (High) 6.3 Meta 2026-08-10

Hugging Face — trending familier

#Rod-repoTrendingLikes DownloadsParamsAlder
1 deepseek-ai/DeepSeek-V4.1-Flash-1017 +3 derivater 880 3887 131062 763.2B 13 d
2 RecursiveSI/flint +1 derivater 480 1246 16598 39.8B 8 d
3 XHToken/Spark-X2.6-4B-Base +2 derivater 450 911 10163 4.1B 11 d
4 nvidia/NVIDIA-Nemotron-3.5-Transcribe-30B-A4B 402 420 4276 33.7B 22 d
5 Lightricks/LTX-2.5 389 9873 2269999 99 d
6 baidu/Unlimited-OCR-2 376 414 18141 3.3B 4 d
7 MiniMaxAI/MiniMax-M3.1-Flash 352 491 15738 220.5B 13 d
8 mistralai/Mistral-Large-4-1T-A52B 311 1234 278 29 d
9 Qwen/Qwen3.8-27B +9 derivater 307 18693 7509924 27.8B 86 d
10 Qwen/Qwen3.5-4B-Base +2 derivater 271 287 12237 4.5B 5 d
11 nvidia/NVIDIA-Nemotron-3.5-Bolt-E4B-BF16 +2 derivater 242 259 57228 25 d
12 LiquidAI/LFM3-Omni-2B +1 derivater 206 211 1814 2.2B 8 d
13 openbmb/MiniCPM-OCR-1.0 196 331 1597 1.3B 8 d
14 facebook/sam3 178 5438 1884719 859.9M 357 d
15 HuggingFaceBio/Carbon-X 177 271 2808 14.8B 9 d

Params er safetensors.total, altså lagrede vægte inklusive kvantiserede tensorer — ikke et mål for hvad modellen kræver af hardware.