New AI and Super Intelligence Model Releases
Newest releases first. A new model goes live as soon as its identity and a primary source are confirmed — provisional listings fill in as benchmark stewards and the provider publish.
Claude Haiku 5.5
Anthropic · SI 64.8 · 100% confidence 100 percent, Full
Open source ↗ · first seen Oct 8, 2026
Mistral Large 4 provisional
mistral · SI 56.1 · 32% confidence 32 percent, Low
Open source ↗ · first seen Oct 8, 2026
Grok Imagine Video 1.5 Lite provisional
xAI · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Ling 3.1 Flash provisional
inclusionai · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
GPT-6.1 Sol
OpenAI · SI 73.0 · 100% confidence 100 percent, Full
Open source ↗ · first seen Oct 8, 2026
Claude Sonnet 5.5
Anthropic · SI 69.4 · 93% confidence 93 percent, High
Open source ↗ · first seen Oct 8, 2026
MiniMax-M3.1-Flash-Preview provisional
minimax · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
LongCat-2.5-Preview provisional
meituan · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Qwen 3.8 Max Prime provisional
Alibaba / Qwen · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Claude Opus 5.5
Anthropic · SI 78.4 · 100% confidence 100 percent, Full
Open source ↗ · first seen Oct 8, 2026
Gemini 3.1 Flash TTS Preview provisional
Google · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
GPT-6 Luna
OpenAI · SI 61.6 · 100% confidence 100 percent, Full
Open source ↗ · first seen Oct 8, 2026
GPT-6 Sol
OpenAI · SI 68.9 · 100% confidence 100 percent, Full
Open source ↗ · first seen Oct 8, 2026
MiMo-V2.6-Flash provisional open
xiaomi · SI 55.8 · 88% confidence 88 percent, High
Open source ↗ · first seen Oct 8, 2026
MiMo-V2.6-Pro provisional open
xiaomi · SI 57.4 · 88% confidence 88 percent, High
Open source ↗ · first seen Oct 8, 2026
Viv Fast provisional
vivgrid · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Grok 4.7
xAI · SI 65.5 · 100% confidence 100 percent, Full
Open source ↗ · first seen Oct 8, 2026
MiMo-V2.6-Pro-UltraSpeed provisional open
xiaomi · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Qwen3.8 Omni Flash provisional
Alibaba / Qwen · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Pareto provisional
unbiased · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Arrow 2 provisional
quiverai · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Arrow 2 Telos provisional
quiverai · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Step 5 Preview provisional
stepfun · SI 54.1 · 75% confidence 75 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Qwen-Image-2.1 provisional open
Alibaba / Qwen · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Kimi K2.8 Preview provisional
Moonshot AI · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
DeepSeek V4.1 Flash open
DeepSeek · SI 66.2 · 69% confidence 69 percent, Medium
Open source ↗ · first seen Oct 8, 2026
North Small Translate provisional open
cohere · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
GPT Image 2.5 Flare provisional
OpenAI · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
GPT Image 2.5 Sunburst provisional
OpenAI · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
MiniCPM5-2B provisional open
openbmb · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
GPT-6 Astra (Fast) provisional
OpenAI · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
GPT-6 Astra
OpenAI · SI 77.6 · 100% confidence 100 percent, Full
Open source ↗ · first seen Oct 8, 2026
Qwen3.8 Max 0902 provisional
Alibaba / Qwen · SI 51.9 · 40% confidence 40 percent, Low
Open source ↗ · first seen Oct 8, 2026
Gemini 3.8 Flash
Google · SI 68.8 · 100% confidence 100 percent, Full
Open source ↗ · first seen Oct 8, 2026
Muse Spark 1.3 provisional
Meta · SI 70.1 · 93% confidence 93 percent, High
Open source ↗ · first seen Oct 8, 2026
Claude Fable 5.1
Anthropic · SI 80.2 · 100% confidence 100 percent, Full
Open source ↗ · first seen Oct 8, 2026
Hy4 preview provisional open
tencent · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Qwen3.8 Flash Next provisional open
Alibaba / Qwen · SI 54.3 · 21% confidence 21 percent, Low
Open source ↗ · first seen Oct 8, 2026
Ling 3.0 Flash Fin provisional
inclusionai · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Qwen3.8 Flash provisional
Alibaba / Qwen · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Gemini 3.5 Transcribe Live provisional
Google · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
GLM-5.3-Flash provisional open
Z.ai · SI 66.2 · 100% confidence 100 percent, Full
Open source ↗ · first seen Oct 8, 2026
DeepSeek V4 Flash Vision Exp provisional open
DeepSeek · SI 54.8 · 21% confidence 21 percent, Low
Open source ↗ · first seen Oct 8, 2026
Ornith 1.5 35B A3B provisional open
deepreinforce · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Qwen3.8 27B provisional open
Alibaba / Qwen · SI 61.9 · 69% confidence 69 percent, Medium
Open source ↗ · first seen Oct 8, 2026
GLM-5.3 provisional open
Z.ai · SI 67.7 · 93% confidence 93 percent, High
Open source ↗ · first seen Oct 8, 2026
Gemini 3.7 Flash
Google · SI 70.0 · 100% confidence 100 percent, Full
Open source ↗ · first seen Oct 8, 2026
Toast 1 provisional
mixedbread · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Qwen3.8 2.4T A95B provisional open
Alibaba / Qwen · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
DeepSeek V4 Pro 0813 provisional open
DeepSeek · SI 63.0 · 69% confidence 69 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Grok 4.6 provisional
xAI · SI 66.6 · 100% confidence 100 percent, Full
Open source ↗ · first seen Oct 8, 2026
MAI-Code-1.1-Flash provisional
microsoft · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Nemotron 3.5 Lightning 30B A3B provisional open
nvidia · SI 46.1 · 0% confidence 0 percent, Low
Open source ↗ · first seen Oct 8, 2026
Muse Glimmer 30B provisional open
Meta · SI 50.9 · 6% confidence 6 percent, Low
Open source ↗ · first seen Oct 8, 2026
Motif 3 provisional open
motif-technologies · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
GPT-5.6 Cyber provisional
OpenAI · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Grok Imagine Image 2.0 provisional
xAI · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Solar Pro 4 provisional
upstage · SI 53.3 · 0% confidence 0 percent, Low
Open source ↗ · first seen Oct 8, 2026
Muse Spark 1.2 provisional
Meta · SI 62.3 · 53% confidence 53 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Grok Voice STT 1.0 provisional
xAI · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Qwen3.8 Max provisional
Alibaba / Qwen · SI 62.0 · 53% confidence 53 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Sakana Namazu provisional
sakana · SI 53.5 · 0% confidence 0 percent, Low
Open source ↗ · first seen Oct 8, 2026
Seedance 2.5 provisional
bytedance-seed · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
DeepSeek V4 Flash 0731 provisional open
DeepSeek · SI 61.0 · 69% confidence 69 percent, Medium
Open source ↗ · first seen Oct 8, 2026
MiniMax H3 provisional open
minimax · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Grok Voice TTS 1.0 provisional
xAI · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Inkling Small provisional open
thinkingmachines · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Claude Opus 5 provisional
Anthropic · SI 74.9 · 100% confidence 100 percent, Full
Open source ↗ · first seen Oct 8, 2026
Gemini 3.5 Flash Lite
Google · SI 55.2 · 100% confidence 100 percent, Full
Open source ↗ · first seen Oct 8, 2026
Gemini 3.6 Flash
Google · SI 65.6 · 100% confidence 100 percent, Full
Open source ↗ · first seen Oct 8, 2026
Gemini Flash-Lite Latest provisional
Google · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Laguna S 2.1 provisional open
poolside · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Qwen3.8 Max Preview provisional
Alibaba / Qwen · SI 50.9 · 6% confidence 6 percent, Low
Open source ↗ · first seen Oct 8, 2026
Kimi K3 provisional open
Moonshot AI · SI 71.0 · 100% confidence 100 percent, Full
Open source ↗ · first seen Oct 8, 2026
Qwen3.7 Flash provisional
Alibaba / Qwen · SI 54.5 · 32% confidence 32 percent, Low
Open source ↗ · first seen Oct 8, 2026
Inkling provisional open
thinkingmachines · SI 66.9 · 0% confidence 0 percent, Low
Open source ↗ · first seen Oct 8, 2026
GPT-5.6 Luna provisional
OpenAI · SI 63.4 · 100% confidence 100 percent, Full
Open source ↗ · first seen Oct 8, 2026
GPT-5.6 Sol provisional
OpenAI · SI 72.1 · 100% confidence 100 percent, Full
Open source ↗ · first seen Oct 8, 2026
GPT-5.6 Terra provisional
OpenAI · SI 68.2 · 100% confidence 100 percent, Full
Open source ↗ · first seen Oct 8, 2026
Grok 4.5 provisional
xAI · SI 64.8 · 100% confidence 100 percent, Full
Open source ↗ · first seen Oct 8, 2026
GPT-Realtime-2.1 provisional
OpenAI · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Hy3 provisional open
tencent · SI 60.7 · 88% confidence 88 percent, High
Open source ↗ · first seen Oct 8, 2026
Laguna XS 2.1 provisional open
poolside · SI 50.5 · 0% confidence 0 percent, Low
Open source ↗ · first seen Oct 8, 2026
Claude Sonnet 5 provisional
Anthropic · SI 67.7 · 93% confidence 93 percent, High
Open source ↗ · first seen Oct 8, 2026
Nano Banana 2 Lite provisional
Google · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Gemini Omni Flash Preview provisional
Google · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
LongCat-2.0 provisional
meituan · SI 56.6 · 0% confidence 0 percent, Low
Open source ↗ · first seen Oct 8, 2026
Ornith 1.0 31B provisional open
deepreinforce · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Ornith 1.0 35B provisional open
deepreinforce · SI 53.5 · 0% confidence 0 percent, Low
Open source ↗ · first seen Oct 8, 2026
Ornith 1.0 397B provisional open
deepreinforce · SI 56.3 · 0% confidence 0 percent, Low
Open source ↗ · first seen Oct 8, 2026
Ornith 1.0 9B provisional open
deepreinforce · SI 50.4 · 0% confidence 0 percent, Low
Open source ↗ · first seen Oct 8, 2026
Seed 2.1 Pro provisional
bytedance-seed · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Seed 2.1 Turbo provisional
bytedance-seed · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Seed Character provisional
bytedance-seed · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Seed Evolving provisional
bytedance-seed · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Fugu provisional
sakana · SI 55.9 · 0% confidence 0 percent, Low
Open source ↗ · first seen Oct 8, 2026
Fugu Ultra provisional
sakana · SI 57.9 · 0% confidence 0 percent, Low
Open source ↗ · first seen Oct 8, 2026
GLM-5.2 provisional open
Z.ai · SI 65.7 · 100% confidence 100 percent, Full
Open source ↗ · first seen Oct 8, 2026
Kimi K2.7 Code provisional open
Moonshot AI · SI 56.3 · 48% confidence 48 percent, Low
Open source ↗ · first seen Oct 8, 2026
Kimi K2.7 Code Highspeed provisional open
Moonshot AI · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Claude Fable 5 provisional
Anthropic · SI 76.8 · 100% confidence 100 percent, Full
Open source ↗ · first seen Oct 8, 2026
Claude Mythos 5 provisional
Anthropic · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
North Mini Code provisional open
cohere · SI 50.4 · 13% confidence 13 percent, Low
Open source ↗ · first seen Oct 8, 2026
DiffusionGemma 26B A4B IT provisional open
Google · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Gemini 3.5 Live Translate Preview provisional
Google · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Gemma 4 12B IT provisional open
Google · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
MiMo-V2.5-Pro-UltraSpeed provisional open
xiaomi · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Nemotron 3 Ultra 550B A55B provisional open
nvidia · SI 61.2 · 0% confidence 0 percent, Low
Open source ↗ · first seen Oct 8, 2026
Nemotron 3.5 Content Safety provisional open
nvidia · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Qwen3.7 Plus provisional
Alibaba / Qwen · SI 62.7 · 64% confidence 64 percent, Medium
Open source ↗ · first seen Oct 8, 2026
MAI-Code-1-Flash provisional
microsoft · SI 50.1 · 0% confidence 0 percent, Low
Open source ↗ · first seen Oct 8, 2026
Nex-N2-Pro provisional open
nex-agi · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
MiniMax-M3 provisional open
minimax · SI 68.0 · 100% confidence 100 percent, Full
Open source ↗ · first seen Oct 8, 2026
Grok Imagine Video 1.5 provisional
xAI · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Step 3.7 Flash provisional open
stepfun · SI 50.3 · 13% confidence 13 percent, Low
Open source ↗ · first seen Oct 8, 2026
Claude Opus 4.8 provisional
Anthropic · SI 67.5 · 100% confidence 100 percent, Full
Open source ↗ · first seen Oct 8, 2026
Nano Banana Pro provisional
Google · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Nano Banana 2 provisional
Google · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Qwen3.7 Max provisional
Alibaba / Qwen · SI 58.6 · 53% confidence 53 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Command A Plus provisional open
cohere · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Gemini 3.5 Flash
Google · SI 68.3 · 100% confidence 100 percent, Full
Open source ↗ · first seen Oct 8, 2026
Gemini Flash Latest provisional
Google · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
MiniCPM5-1B provisional open
openbmb · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Gemini 3.1 Flash Lite provisional
Google · SI 53.2 · 32% confidence 32 percent, Low
Open source ↗ · first seen Oct 8, 2026
GPT Realtime Whisper provisional
OpenAI · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
GPT-5.5 Instant provisional
OpenAI · SI 58.3 · 64% confidence 64 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Mistral Medium 3.5 provisional open
mistral · SI 58.1 · 37% confidence 37 percent, Low
Open source ↗ · first seen Oct 8, 2026
Mistral Medium (latest) provisional open
mistral · SI 51.2 · 6% confidence 6 percent, Low
Open source ↗ · first seen Oct 8, 2026
Nemotron 3 Nano Omni 30B A3B Reasoning provisional open
nvidia · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Laguna M.1 provisional open
poolside · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Laguna XS.2 provisional open
poolside · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Qwen3.6 Flash provisional
Alibaba / Qwen · SI 54.9 · 32% confidence 32 percent, Low
Open source ↗ · first seen Oct 8, 2026
DeepSeek V4 Flash provisional open
DeepSeek · SI 56.2 · 53% confidence 53 percent, Medium
Open source ↗ · first seen Oct 8, 2026
DeepSeek V4 Pro provisional open
DeepSeek · SI 63.1 · 85% confidence 85 percent, High
Open source ↗ · first seen Oct 8, 2026
GPT-5.5
OpenAI · SI 69.5 · 100% confidence 100 percent, Full
Open source ↗ · first seen Oct 8, 2026
GPT-5.5 Pro
OpenAI · SI 63.5 · 53% confidence 53 percent, Medium
Open source ↗ · first seen Oct 8, 2026
DeepSeek V4 Flash 0423 provisional open
DeepSeek · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
DeepSeek V4 Pro 0423 provisional open
DeepSeek · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Qwen3.6 27B provisional open
Alibaba / Qwen · SI 55.6 · 53% confidence 53 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Gemini Embedding 2 provisional
Google · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
MiMo-V2.5 provisional open
xiaomi · SI 56.4 · 88% confidence 88 percent, High
Open source ↗ · first seen Oct 8, 2026
MiMo-V2.5-Pro provisional open
xiaomi · SI 62.6 · 88% confidence 88 percent, High
Open source ↗ · first seen Oct 8, 2026
Deep Research Max Preview provisional
Google · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Gemini Deep Research Preview provisional
Google · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Kimi K2.6 provisional open
Moonshot AI · SI 64.0 · 85% confidence 85 percent, High
Open source ↗ · first seen Oct 8, 2026
GPT-Image-2 provisional
OpenAI · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Qwen3.6 Max Preview provisional
Alibaba / Qwen · SI 66.4 · 64% confidence 64 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Hy3 preview provisional open
tencent · SI 52.1 · 13% confidence 13 percent, Low
Open source ↗ · first seen Oct 8, 2026
Qwen3.6 35B-A3B provisional open
Alibaba / Qwen · SI 56.3 · 37% confidence 37 percent, Low
Open source ↗ · first seen Oct 8, 2026
Grok 4.3 provisional
xAI · SI 54.6 · 80% confidence 80 percent, High
Open source ↗ · first seen Oct 8, 2026
Claude Opus 4.7 provisional
Anthropic · SI 70.1 · 100% confidence 100 percent, Full
Open source ↗ · first seen Oct 8, 2026
Nemotron 3 Content Safety provisional open
nvidia · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Grok Build 0.1 provisional
xAI · SI 52.0 · 16% confidence 16 percent, Low
Open source ↗ · first seen Oct 8, 2026
Gemini Robotics-ER 1.6 Preview provisional
Google · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Muse Spark 1.1 provisional
Meta · SI 61.2 · 53% confidence 53 percent, Medium
Open source ↗ · first seen Oct 8, 2026
GLM-5.1 provisional open
Z.ai · SI 63.3 · 64% confidence 64 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Grok Imagine Image Quality provisional
xAI · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Qwen3.6 Plus provisional
Alibaba / Qwen · SI 61.6 · 80% confidence 80 percent, High
Open source ↗ · first seen Oct 8, 2026
Gemma 4 26B A4B IT provisional open
Google · SI 63.4 · 64% confidence 64 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Gemma 4 31B IT provisional open
Google · SI 63.0 · 64% confidence 64 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Gemma 4 E2B IT provisional open
Google · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Gemma 4 E4B IT provisional open
Google · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Step 3.5 Flash 2603 provisional open
stepfun · SI 48.5 · 13% confidence 13 percent, Low
Open source ↗ · first seen Oct 8, 2026
Trinity Large Thinking provisional open
arcee-ai · SI 52.5 · 0% confidence 0 percent, Low
Open source ↗ · first seen Oct 8, 2026
GLM-5V-Turbo provisional
Z.ai · SI 54.0 · 32% confidence 32 percent, Low
Open source ↗ · first seen Oct 8, 2026
Veo 3.1 Lite Preview provisional
Google · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Llama Nemotron Rerank VL 1B v2 provisional open
nvidia · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Gemini 3.1 Flash Live Preview provisional
Google · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Lyria 3 Clip Preview provisional
Google · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Lyria 3 Pro Preview provisional
Google · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Nemotron Cascade 2 30B A3B provisional open
nvidia · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
MiniMax-M2.7 provisional open
minimax · SI 57.9 · 88% confidence 88 percent, High
Open source ↗ · first seen Oct 8, 2026
MiniMax-M2.7-highspeed provisional open
minimax · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
MiMo-V2-Omni provisional
xiaomi · SI 53.8 · 75% confidence 75 percent, Medium
Open source ↗ · first seen Oct 8, 2026
MiMo-V2-Pro provisional
xiaomi · SI 54.0 · 75% confidence 75 percent, Medium
Open source ↗ · first seen Oct 8, 2026
GPT-5.4 mini
OpenAI · SI 55.8 · 69% confidence 69 percent, Medium
Open source ↗ · first seen Oct 8, 2026
GPT-5.4 nano
OpenAI · SI 56.6 · 69% confidence 69 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Mistral Small 4 provisional open
mistral · SI 48.5 · 6% confidence 6 percent, Low
Open source ↗ · first seen Oct 8, 2026
Mistral Small (latest) provisional open
mistral · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Nemotron VoiceChat provisional open
nvidia · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
GLM-5-Turbo provisional
Z.ai · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Nemotron 3 Super 120B A12B provisional open
nvidia · SI 53.1 · 0% confidence 0 percent, Low
Open source ↗ · first seen Oct 8, 2026
Grok 4.20 (Non-Reasoning) provisional
xAI · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Grok 4.20 (Reasoning) provisional
xAI · SI 57.7 · 80% confidence 80 percent, High
Open source ↗ · first seen Oct 8, 2026
GPT-5.4
OpenAI · SI 61.9 · 69% confidence 69 percent, Medium
Open source ↗ · first seen Oct 8, 2026
GPT-5.4 Pro provisional
OpenAI · SI 53.8 · 21% confidence 21 percent, Low
Open source ↗ · first seen Oct 8, 2026
Gemini 3.1 Flash Lite Preview provisional
Google · SI 42.6 · 48% confidence 48 percent, Low
Open source ↗ · first seen Oct 8, 2026
GPT-5.3 Chat (latest) provisional
OpenAI · SI 53.3 · 32% confidence 32 percent, Low
Open source ↗ · first seen Oct 8, 2026
Nano Banana 2 Preview provisional
Google · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
GPT-5.3 Codex provisional
OpenAI · SI 56.2 · 32% confidence 32 percent, Low
Open source ↗ · first seen Oct 8, 2026
Qwen3.5 122B-A10B provisional open
Alibaba / Qwen · SI 56.8 · 37% confidence 37 percent, Low
Open source ↗ · first seen Oct 8, 2026
Qwen3.5 27B provisional open
Alibaba / Qwen · SI 56.8 · 37% confidence 37 percent, Low
Open source ↗ · first seen Oct 8, 2026
Qwen3.5 35B-A3B provisional open
Alibaba / Qwen · SI 62.7 · 64% confidence 64 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Qwen3.5 9B provisional open
Alibaba / Qwen · SI 55.1 · 32% confidence 32 percent, Low
Open source ↗ · first seen Oct 8, 2026
Qwen3.5 Flash provisional
Alibaba / Qwen · SI 59.5 · 64% confidence 64 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Gemini 3.1 Pro Preview provisional
Google · SI 68.1 · 100% confidence 100 percent, Full
Open source ↗ · first seen Oct 8, 2026
Gemini 3.1 Pro Preview Custom Tools provisional
Google · SI 56.4 · 32% confidence 32 percent, Low
Open source ↗ · first seen Oct 8, 2026
Trendyol Asure 12B provisional open
trendyol · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Sarvam 30B provisional open
sarvam · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Claude Sonnet 4.6 provisional
Anthropic · SI 68.2 · 100% confidence 100 percent, Full
Open source ↗ · first seen Oct 8, 2026
Tiny Aya Earth provisional open
cohere · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Tiny Aya Fire provisional open
cohere · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Tiny Aya Global provisional open
cohere · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Tiny Aya Water provisional open
cohere · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Qwen3.5 Plus provisional
Alibaba / Qwen · SI 59.0 · 32% confidence 32 percent, Low
Open source ↗ · first seen Oct 8, 2026
Qwen3.5 397B-A17B provisional open
Alibaba / Qwen · SI 64.0 · 69% confidence 69 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Seed 2.0 Code provisional
bytedance-seed · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Seed 2.0 Lite provisional
bytedance-seed · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Seed 2.0 Mini provisional
bytedance-seed · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Seed 2.0 Pro provisional
bytedance-seed · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
MiniMax-M2.5-highspeed provisional open
minimax · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Seedance 2.0 provisional
bytedance-seed · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
MiniMax-M2.5 provisional open
minimax · SI 56.7 · 88% confidence 88 percent, High
Open source ↗ · first seen Oct 8, 2026
GLM-5 provisional open
Z.ai · SI 58.9 · 85% confidence 85 percent, High
Open source ↗ · first seen Oct 8, 2026
Llama Nemotron Embed VL 1B v2 provisional open
nvidia · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Claude Opus 4.6 provisional
Anthropic · SI 70.5 · 100% confidence 100 percent, Full
Open source ↗ · first seen Oct 8, 2026
GPT-5.3 Codex Spark provisional
OpenAI · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Qwen3 Coder Next provisional open
Alibaba / Qwen · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Step 3.5 Flash provisional open
stepfun · SI 53.0 · 88% confidence 88 percent, High
Open source ↗ · first seen Oct 8, 2026
Trinity Large Preview provisional open
arcee-ai · SI 52.5 · 0% confidence 0 percent, Low
Open source ↗ · first seen Oct 8, 2026
DeepSeek OCR 2 provisional open
DeepSeek · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
MiniMax-M2 Her provisional
minimax · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Nemotron Content Safety Reasoning 4B provisional open
nvidia · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
GLM-4.7-Flash provisional open
Z.ai · SI 53.3 · 69% confidence 69 percent, Medium
Open source ↗ · first seen Oct 8, 2026
GLM-4.7-FlashX provisional open
Z.ai · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
GLM-Image provisional open
Z.ai · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
GPT-5.2 Codex provisional
OpenAI · SI 54.0 · 21% confidence 21 percent, Low
Open source ↗ · first seen Oct 8, 2026
Jamba Mini provisional open
ai21 · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Kimi K2.5 provisional open
Moonshot AI · SI 58.8 · 100% confidence 100 percent, Full
Open source ↗ · first seen Oct 8, 2026
Solar Pro 3 provisional
upstage · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Seed 1.8 provisional
bytedance-seed · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
MiniMax-M2.1 provisional open
minimax · SI 50.6 · 13% confidence 13 percent, Low
Open source ↗ · first seen Oct 8, 2026
GLM-4.7 provisional open
Z.ai · SI 61.3 · 69% confidence 69 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Gemini 3 Flash Preview provisional
Google · SI 54.4 · 85% confidence 85 percent, High
Open source ↗ · first seen Oct 8, 2026
GPT-Image-1.5 provisional
OpenAI · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
MiMo-V2-Flash provisional open
xiaomi · SI 53.4 · 75% confidence 75 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Nemotron 3 Nano 30B A3B provisional open
nvidia · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
GPT-5.2
OpenAI · SI 51.4 · 53% confidence 53 percent, Medium
Open source ↗ · first seen Oct 8, 2026
GPT-5.2 Chat provisional
OpenAI · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
GPT-5.2 Pro provisional
OpenAI · SI 52.3 · 16% confidence 16 percent, Low
Open source ↗ · first seen Oct 8, 2026
Devstral 2 provisional open
mistral · SI 49.5 · 6% confidence 6 percent, Low
Open source ↗ · first seen Oct 8, 2026
Devstral Small 2 provisional open
mistral · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
GLM-4.6V provisional open
Z.ai · SI 53.1 · 32% confidence 32 percent, Low
Open source ↗ · first seen Oct 8, 2026
GLM-4.6V-Flash provisional open
Z.ai · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
GPT-5.1 Codex Max provisional
OpenAI · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Nova 2 Lite provisional
amazon · SI 52.9 · 0% confidence 0 percent, Low
Open source ↗ · first seen Oct 8, 2026
Devstral 2 (latest) provisional open
mistral · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Ministral 14B provisional open
mistral · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Ministral 3 14B provisional open
mistral · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Ministral 3 3B provisional open
mistral · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Ministral 3 8B provisional open
mistral · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Mistral Large 3 provisional open
mistral · SI 44.7 · 37% confidence 37 percent, Low
Open source ↗ · first seen Oct 8, 2026
Trinity Mini provisional open
arcee-ai · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Trinity Nano Preview provisional open
arcee-ai · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
DeepSeek Chat provisional open
DeepSeek · SI 53.5 · 32% confidence 32 percent, Low
Open source ↗ · first seen Oct 8, 2026
DeepSeek Reasoner provisional open
DeepSeek · SI 58.9 · 32% confidence 32 percent, Low
Open source ↗ · first seen Oct 8, 2026
DeepSeek V3.2 provisional open
DeepSeek · SI 52.8 · 48% confidence 48 percent, Low
Open source ↗ · first seen Oct 8, 2026
Claude Opus 4.5 (latest) provisional
Anthropic · SI 50.0 · 0% confidence 0 percent, Low
Open source ↗ · first seen Oct 8, 2026
Nano Banana Pro Preview provisional
Google · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Grok 4.1 Fast provisional
xAI · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Grok 4.1 Fast (Reasoning) provisional
xAI · SI 53.6 · 32% confidence 32 percent, Low
Open source ↗ · first seen Oct 8, 2026
Gemini 3 Pro Preview provisional
Google · SI 56.8 · 53% confidence 53 percent, Medium
Open source ↗ · first seen Oct 8, 2026
GPT-5.1 provisional
OpenAI · SI 51.4 · 48% confidence 48 percent, Low
Open source ↗ · first seen Oct 8, 2026
GPT-5.1 Chat provisional
OpenAI · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
GPT-5.1 Codex provisional
OpenAI · SI 50.0 · 0% confidence 0 percent, Low
Open source ↗ · first seen Oct 8, 2026
GPT-5.1 Codex mini provisional
OpenAI · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Kimi K2 Thinking provisional open
Moonshot AI · SI 50.7 · 6% confidence 6 percent, Low
Open source ↗ · first seen Oct 8, 2026
Kimi K2 Thinking Turbo provisional open
Moonshot AI · SI 66.0 · 64% confidence 64 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Claude Opus 4.5 provisional
Anthropic · SI 64.6 · 100% confidence 100 percent, Full
Open source ↗ · first seen Oct 8, 2026
GPT OSS Safeguard 120B provisional open
OpenAI · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
GPT OSS Safeguard 20B provisional open
OpenAI · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Llama 3.1 Nemotron Safety Guard 8B v3 provisional open
nvidia · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Nemotron Nano 12B v2 VL provisional open
nvidia · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
MiniMax-M2 provisional open
minimax · SI 55.7 · 88% confidence 88 percent, High
Open source ↗ · first seen Oct 8, 2026
Claude Haiku 4.5 (latest) provisional
Anthropic · SI 49.5 · 6% confidence 6 percent, Low
Open source ↗ · first seen Oct 8, 2026
Claude Haiku 4.5 provisional
Anthropic · SI 58.1 · 64% confidence 64 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Seed 1.6 provisional
bytedance-seed · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Veo 3.1 Fast Preview provisional
Google · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Veo 3.1 Preview provisional
Google · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Gemini 2.5 Computer Use Preview provisional
Google · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
GPT-5 Pro provisional
OpenAI · SI 49.3 · 16% confidence 16 percent, Low
Open source ↗ · first seen Oct 8, 2026
Granite-4.0-H-Micro provisional open
ibm · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Granite-4.0-H-Small provisional open
ibm · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Gemini 2.5 Flash TTS provisional
Google · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Gemini 2.5 Pro TTS provisional
Google · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
GLM-4.6 provisional open
Z.ai · SI 48.5 · 37% confidence 37 percent, Low
Open source ↗ · first seen Oct 8, 2026
Claude Sonnet 4.5 (latest) provisional
Anthropic · SI 50.4 · 6% confidence 6 percent, Low
Open source ↗ · first seen Oct 8, 2026
Claude Sonnet 4.5 provisional
Anthropic · SI 52.7 · 80% confidence 80 percent, High
Open source ↗ · first seen Oct 8, 2026
Gemma-SEA-LION-v4-27B-IT provisional open
aisingapore · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Qwen3 Max provisional
Alibaba / Qwen · SI 48.7 · 6% confidence 6 percent, Low
Open source ↗ · first seen Oct 8, 2026
Qwen3 VL 235B A22B Instruct provisional open
Alibaba / Qwen · SI 53.8 · 32% confidence 32 percent, Low
Open source ↗ · first seen Oct 8, 2026
Qwen3 VL 235B A22B Thinking provisional open
Alibaba / Qwen · SI 53.5 · 32% confidence 32 percent, Low
Open source ↗ · first seen Oct 8, 2026
Qwen3-VL Plus provisional
Alibaba / Qwen · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Magistral Small 1.2 provisional open
mistral · SI 47.7 · 32% confidence 32 percent, Low
Open source ↗ · first seen Oct 8, 2026
GPT-5-Codex provisional
OpenAI · SI 49.5 · 6% confidence 6 percent, Low
Open source ↗ · first seen Oct 8, 2026
Apertus 70B provisional open
swiss-ai · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Apertus 8B provisional open
swiss-ai · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Sarvam 105B provisional open
sarvam · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Qwen3-Next 80B-A3B Instruct provisional open
Alibaba / Qwen · SI 53.7 · 32% confidence 32 percent, Low
Open source ↗ · first seen Oct 8, 2026
Qwen3-Next 80B-A3B (Thinking) provisional open
Alibaba / Qwen · SI 53.0 · 32% confidence 32 percent, Low
Open source ↗ · first seen Oct 8, 2026
Seed 1.6 Flash provisional
bytedance-seed · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Command A Translate provisional open
cohere · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Nano Banana provisional
Google · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Command A Reasoning provisional open
cohere · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
DeepSeek-V3.1 provisional open
DeepSeek · SI 53.7 · 32% confidence 32 percent, Low
Open source ↗ · first seen Oct 8, 2026
Gemini 2.0 Flash provisional
Google · SI 44.7 · 16% confidence 16 percent, Low
Open source ↗ · first seen Oct 8, 2026
Nemotron Nano 9B v2 provisional open
nvidia · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Seed 1.6 Vision provisional
bytedance-seed · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
GLM-4.5V provisional open
Z.ai · SI 41.6 · 37% confidence 37 percent, Low
Open source ↗ · first seen Oct 8, 2026
GPT-5
OpenAI · SI 50.9 · 53% confidence 53 percent, Medium
Open source ↗ · first seen Oct 8, 2026
GPT-5 Chat (latest) provisional
OpenAI · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
GPT-5 Mini provisional
OpenAI · SI 49.1 · 48% confidence 48 percent, Low
Open source ↗ · first seen Oct 8, 2026
GPT-5 Nano provisional
OpenAI · SI 42.7 · 48% confidence 48 percent, Low
Open source ↗ · first seen Oct 8, 2026
Claude Opus 4.1 (latest) provisional
Anthropic · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Claude Opus 4.1 provisional
Anthropic · SI 53.5 · 80% confidence 80 percent, High
Open source ↗ · first seen Oct 8, 2026
GPT OSS 120B provisional open
OpenAI · SI 53.9 · 64% confidence 64 percent, Medium
Open source ↗ · first seen Oct 8, 2026
GPT OSS 20B provisional open
OpenAI · SI 52.5 · 64% confidence 64 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Command A Vision provisional open
cohere · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Codestral 25.08 provisional
mistral · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Qwen Flash provisional
Alibaba / Qwen · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Qwen3 Coder Flash provisional
Alibaba / Qwen · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
GLM-4.5 provisional open
Z.ai · SI 50.8 · 37% confidence 37 percent, Low
Open source ↗ · first seen Oct 8, 2026
GLM-4.5-Air provisional open
Z.ai · SI 45.3 · 37% confidence 37 percent, Low
Open source ↗ · first seen Oct 8, 2026
GLM-4.5-Flash provisional
Z.ai · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Llama 3.3 Nemotron Super 49B v1.5 provisional open
nvidia · SI 52.5 · 0% confidence 0 percent, Low
Open source ↗ · first seen Oct 8, 2026
Qwen3 Coder Plus provisional
Alibaba / Qwen · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Gemini 2.5 Flash-Lite provisional
Google · SI 48.0 · 6% confidence 6 percent, Low
Open source ↗ · first seen Oct 8, 2026
Qwen3 235B-A22B Instruct 2507 provisional open
Alibaba / Qwen · SI 42.4 · 48% confidence 48 percent, Low
Open source ↗ · first seen Oct 8, 2026
Voxtral Mini 3B 2507 provisional open
mistral · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Voxtral Small 24B 2507 provisional open
mistral · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Voxtral Small (latest) provisional open
mistral · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Gemini Embedding 001 provisional
Google · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Devstral Medium provisional
mistral · SI 50.5 · 6% confidence 6 percent, Low
Open source ↗ · first seen Oct 8, 2026
Devstral Small provisional open
mistral · SI 49.8 · 6% confidence 6 percent, Low
Open source ↗ · first seen Oct 8, 2026
Jamba Large provisional open
ai21 · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Mistral Small 3.2 provisional open
mistral · SI 52.5 · 32% confidence 32 percent, Low
Open source ↗ · first seen Oct 8, 2026
Gemini 2.5 Flash provisional
Google · SI 46.0 · 37% confidence 37 percent, Low
Open source ↗ · first seen Oct 8, 2026
Mistral Nemotron provisional open
nvidia · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Magistral Small provisional open
mistral · SI 37.1 · 48% confidence 48 percent, Low
Open source ↗ · first seen Oct 8, 2026
o3-pro provisional
OpenAI · SI 51.4 · 16% confidence 16 percent, Low
Open source ↗ · first seen Oct 8, 2026
Gemini 2.5 Pro
Google · SI 45.5 · 85% confidence 85 percent, High
Open source ↗ · first seen Oct 8, 2026
Claude Opus 4 (latest) provisional
Anthropic · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Claude Opus 4 provisional
Anthropic · SI 54.8 · 96% confidence 96 percent, High
Open source ↗ · first seen Oct 8, 2026
Claude Sonnet 4 (latest) provisional
Anthropic · SI 49.7 · 6% confidence 6 percent, Low
Open source ↗ · first seen Oct 8, 2026
Claude Sonnet 4 provisional
Anthropic · SI 45.2 · 96% confidence 96 percent, High
Open source ↗ · first seen Oct 8, 2026
Solar Pro 2 provisional
upstage · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Mistral Medium 3 provisional
mistral · SI 33.0 · 85% confidence 85 percent, High
Open source ↗ · first seen Oct 8, 2026
Nova Premier provisional
amazon · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Qwen3 30B A3B provisional open
Alibaba / Qwen · SI 52.6 · 64% confidence 64 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Palmyra X5 provisional
writer · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
GPT-Image-1 provisional
OpenAI · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
o3 provisional
OpenAI · SI 50.1 · 16% confidence 16 percent, Low
Open source ↗ · first seen Oct 8, 2026
o4-mini provisional
OpenAI · SI 49.6 · 16% confidence 16 percent, Low
Open source ↗ · first seen Oct 8, 2026
Llama 3.1 Nemotron 70B Instruct provisional open
nvidia · SI 51.5 · 0% confidence 0 percent, Low
Open source ↗ · first seen Oct 8, 2026
GPT-4.1 provisional
OpenAI · SI 44.5 · 32% confidence 32 percent, Low
Open source ↗ · first seen Oct 8, 2026
GPT-4.1 mini provisional
OpenAI · SI 45.8 · 16% confidence 16 percent, Low
Open source ↗ · first seen Oct 8, 2026
GPT-4.1 nano provisional
OpenAI · SI 44.3 · 16% confidence 16 percent, Low
Open source ↗ · first seen Oct 8, 2026
Pixtral Large (25.02) provisional
mistral · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Llama 3.1 Nemotron Ultra 253B provisional open
nvidia · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Llama 3.3 Nemotron Super 49B v1 provisional open
nvidia · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Llama 4 Maverick 17B Instruct provisional open
Meta · SI 32.2 · 69% confidence 69 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Llama 4 Scout 17B Instruct provisional open
Meta · SI 51.0 · 64% confidence 64 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Qwen3 235B-A22B provisional open
Alibaba / Qwen · SI 50.3 · 69% confidence 69 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Qwen3 32B provisional open
Alibaba / Qwen · SI 50.4 · 64% confidence 64 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Qwen3-Coder 30B-A3B Instruct provisional open
Alibaba / Qwen · SI 49.4 · 6% confidence 6 percent, Low
Open source ↗ · first seen Oct 8, 2026
Qwen3-Coder 480B-A35B Instruct provisional open
Alibaba / Qwen · SI 52.1 · 37% confidence 37 percent, Low
Open source ↗ · first seen Oct 8, 2026
DeepSeek V3 0324 provisional open
DeepSeek · SI 55.0 · 64% confidence 64 percent, Medium
Open source ↗ · first seen Oct 8, 2026
o1-pro provisional
OpenAI · SI 44.7 · 16% confidence 16 percent, Low
Open source ↗ · first seen Oct 8, 2026
Magistral Medium (latest) provisional
mistral · SI 44.1 · 16% confidence 16 percent, Low
Open source ↗ · first seen Oct 8, 2026
Mistral Small 3.1 24B provisional open
mistral · SI 51.4 · 32% confidence 32 percent, Low
Open source ↗ · first seen Oct 8, 2026
Command A provisional open
cohere · SI 52.4 · 75% confidence 75 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Gemma 3 12B IT provisional open
Google · SI 45.1 · 64% confidence 64 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Gemma 3 27B IT provisional open
Google · SI 43.1 · 64% confidence 64 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Gemma 3 4B IT provisional open
Google · SI 39.2 · 64% confidence 64 percent, Medium
Open source ↗ · first seen Oct 8, 2026
QwQ 32B provisional open
Alibaba / Qwen · SI 48.3 · 64% confidence 64 percent, Medium
Open source ↗ · first seen Oct 8, 2026
QwQ Plus provisional
Alibaba / Qwen · SI 53.9 · 32% confidence 32 percent, Low
Open source ↗ · first seen Oct 8, 2026
Aya Vision 32B provisional open
cohere · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Aya Vision 8B provisional open
cohere · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Command R7B Arabic provisional open
cohere · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Gemini 2.0 Flash-Lite provisional
Google · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Claude Sonnet 3.7 provisional
Anthropic · SI 49.8 · 80% confidence 80 percent, High
Open source ↗ · first seen Oct 8, 2026
MiniMax image-01 provisional
minimax · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Sonar Deep Research provisional
perplexity · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
ALLaM-2-7b provisional open
sdaia · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
DeepSeek-R1 provisional open
DeepSeek · SI 48.0 · 85% confidence 85 percent, High
Open source ↗ · first seen Oct 8, 2026
DeepSeek-R1-Distill-Qwen-32B provisional open
DeepSeek · SI 52.9 · 32% confidence 32 percent, Low
Open source ↗ · first seen Oct 8, 2026
Qwen-Omni Turbo provisional
Alibaba / Qwen · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
DeepSeek-V3 provisional open
DeepSeek · SI 54.1 · 75% confidence 75 percent, Medium
Open source ↗ · first seen Oct 8, 2026
o3-mini provisional
OpenAI · SI 43.5 · 75% confidence 75 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Phi-4-mini provisional open
microsoft · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Llama-3.3-70B-Instruct provisional open
Meta · SI 35.1 · 88% confidence 88 percent, High
Open source ↗ · first seen Oct 8, 2026
o1 provisional
OpenAI · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Nova Lite provisional
amazon · SI 36.6 · 0% confidence 0 percent, Low
Open source ↗ · first seen Oct 8, 2026
Nova Micro provisional
amazon · SI 50.2 · 0% confidence 0 percent, Low
Open source ↗ · first seen Oct 8, 2026
Nova Pro provisional
amazon · SI 37.4 · 0% confidence 0 percent, Low
Open source ↗ · first seen Oct 8, 2026
Command R7B provisional open
cohere · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
GPT-4o (2024-11-20) provisional
OpenAI · SI 48.3 · 13% confidence 13 percent, Low
Open source ↗ · first seen Oct 8, 2026
Mistral Large 2.1 provisional open
mistral · SI 37.4 · 88% confidence 88 percent, High
Open source ↗ · first seen Oct 8, 2026
Qwen2.5-Coder-0.5B provisional open
Alibaba / Qwen · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Qwen2.5-Coder-32B-Instruct provisional open
Alibaba / Qwen · SI 39.6 · 75% confidence 75 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Qwen Turbo provisional
Alibaba / Qwen · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Mistral Large (latest) provisional open
mistral · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Pixtral Large (latest) provisional open
mistral · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Aya Expanse 32B provisional open
cohere · SI 50.4 · 75% confidence 75 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Aya Expanse 8B provisional open
cohere · SI 49.7 · 75% confidence 75 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Claude Haiku 3.5 provisional
Anthropic · SI 41.7 · 75% confidence 75 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Claude Sonnet 3.5 v2 provisional
Anthropic · SI 51.3 · 75% confidence 75 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Ministral 3B provisional open
mistral · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Ministral 8B Instruct provisional open
mistral · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Palmyra X4 provisional
writer · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Whisper 3 Large provisional open
OpenAI · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Whisper Large v3 Turbo provisional open
OpenAI · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Llama-3.2-11B-Vision-Instruct provisional open
Meta · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Llama-3.2-1B provisional open
Meta · SI 32.2 · 75% confidence 75 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Llama-3.2-3B provisional open
Meta · SI 48.3 · 75% confidence 75 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Pixtral 12B provisional open
mistral · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Qwen2.5-VL 72B Instruct provisional open
Alibaba / Qwen · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Command R provisional open
cohere · SI 49.8 · 75% confidence 75 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Command R+ provisional open
cohere · SI 50.5 · 75% confidence 75 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Nemotron Mini 4B Instruct provisional open
nvidia · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
GPT-4o (2024-08-06) provisional
OpenAI · SI 38.8 · 88% confidence 88 percent, High
Open source ↗ · first seen Oct 8, 2026
Llama-3.1-70B-Instruct provisional open
Meta · SI 45.8 · 75% confidence 75 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Llama-3.1-8B-Instruct provisional open
Meta · SI 37.8 · 75% confidence 75 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Llama-Guard-3-8B provisional open
Meta · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
GPT-4o mini provisional
OpenAI · SI 50.0 · 0% confidence 0 percent, Low
Open source ↗ · first seen Oct 8, 2026
Mistral Nemo provisional open
mistral · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
o3-deep-research provisional
OpenAI · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
o4-mini-deep-research provisional
OpenAI · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Codestral-22B-v0.1 provisional open
mistral · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Codestral (latest) provisional open
mistral · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Vision Large provisional
vispark · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Vision Medium provisional
vispark · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Vision Small provisional
vispark · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
GPT-4o provisional
OpenAI · SI 50.0 · 0% confidence 0 percent, Low
Open source ↗ · first seen Oct 8, 2026
GPT-4o (2024-05-13) provisional
OpenAI · SI 49.4 · 75% confidence 75 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Qwen-VL Max provisional
Alibaba / Qwen · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Qwen Max provisional
Alibaba / Qwen · SI 53.0 · 75% confidence 75 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Claude Haiku 3 provisional
Anthropic · SI 39.9 · 75% confidence 75 percent, Medium
Open source ↗ · first seen Oct 8, 2026
Qwen Plus provisional
Alibaba / Qwen · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Qwen-VL Plus provisional
Alibaba / Qwen · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Sonar provisional
perplexity · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Sonar Pro provisional
perplexity · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Sonar Reasoning Pro provisional
perplexity · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
Mistral Embed provisional
mistral · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
GPT-3.5-turbo provisional
OpenAI · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
GPT-4 provisional
OpenAI · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026
GPT-4 Turbo provisional
OpenAI · score not yet reported · confidence 0%
Open source ↗ · first seen Oct 8, 2026