Frontier Model Tracker

Crowd-vote ranking vs benchmarked intelligence, plus the releases behind the moves.

updated Sun, 25 Oct 2026 13:30:06 +0000

Crowd-vote vs intelligence — top 20

OpenAI Anthropic Google xAI Other

Changelog — frontier signal

23 Oct 2026

Kimi K3.1 ProEVALUATIONIntelligence Index 55 Kimi K3.1 Pro launches at 4th on the Intelligence Index and will lead open weights modelsARTICLEArtificial Analysis

19 Oct 2026

HHealthBench and HealthBench Professional: Evaluating models in realistic health conversationsARTICLEArtificial Analysis

18 Oct 2026

DeepSeek V4.1 Flash 1017 (Max)EVALUATIONIntelligence Index 45

16 Oct 2026

RReflection has released Beam, their first model launchARTICLEArtificial Analysis

15 Oct 2026

NVIDIA Releases Nemotron 3.5 Super VL and Nemotron 3.5 Bolt 74B A9BARTICLEArtificial Analysis

14 Oct 2026

WWelcome to the era of embodied AI: Introducing AA-Robotics v0.1ARTICLEArtificial Analysis

13 Oct 2026

AArtificial Analysis Decision Index v1.0ARTICLEArtificial Analysis

12 Oct 2026

AAnnouncing AA-3D: a benchmark for AI models building interactive 3D experiencesARTICLEArtificial Analysis AAnnouncing FrontierCode on Artificial Analysis: can AI models write code maintainers would merge?ARTICLEArtificial Analysis MMuse Wonder (Max)EVALUATIONIntelligence Index 52 Muse Wonder and Muse Wonder Full: a material jump in intelligence over previous Meta modelsARTICLEArtificial Analysis MMuse Wonder Full (Max)EVALUATIONIntelligence Index 56

08 Oct 2026

AAnnouncing Harvey LAB-AA v1.1: adding hallucination checks to raise the bar for agentic legal workARTICLEArtificial Analysis IIntroducing trusted-access models to the Artificial Analysis Cyber IndexARTICLEArtificial Analysis

07 Oct 2026

Anthropic has released Claude Haiku 5.5ARTICLEArtificial Analysis

06 Oct 2026

Mistral has released Mistral Large 4, making France home to the most intelligent model outside the US and ChinaARTICLEArtificial Analysis

01 Oct 2026

AAA-AgentPerf-Local: Benchmarking local AI agents on laptops and workstationsARTICLEArtificial Analysis AAnnouncing Artificial Analysis Capability Indices v1.1ARTICLEArtificial Analysis AAnnouncing Artificial Analysis Intelligence Index v4.2ARTICLEArtificial Analysis AAnnouncing the Artificial Analysis Cyber Index AllianceARTICLEArtificial Analysis AAnnouncing the Artificial Analysis Intelligence Index v4.3ARTICLEArtificial Analysis AAnt Group releases finance-focused Ling-3.0-flash-FinARTICLEArtificial Analysis Benchmarking GPT-6 AstraARTICLEArtificial Analysis Benchmarking Grok 4.7ARTICLEArtificial Analysis Claude Fable 5.1 (Adaptive Reasoning, High Effort, Default Fallback)EVALUATIONIntelligence Index 51 Claude Fable 5.1 (Adaptive Reasoning, Low Effort, Default Fallback)EVALUATIONIntelligence Index 47 Claude Fable 5.1 (Adaptive Reasoning, Max Effort, Default Fallback)EVALUATIONIntelligence Index 53 Claude Fable 5.1 (Adaptive Reasoning, Medium Effort, Default Fallback)EVALUATIONIntelligence Index 49 Claude Fable 5.1 (Adaptive Reasoning, Xhigh Effort, Default Fallback)EVALUATIONIntelligence Index 53 Claude Fable 5.1 (High, Default Fallback)EVALUATIONIntelligence Index 51 Claude Fable 5.1 (Low, Default Fallback)EVALUATIONIntelligence Index 47 Claude Fable 5.1 (Max, Default Fallback)EVALUATIONIntelligence Index 53 Claude Fable 5.1 (Medium, Default Fallback)EVALUATIONIntelligence Index 49 Claude Fable 5.1 (Xhigh, Default Fallback)EVALUATIONIntelligence Index 53 Claude Fable 5.1 tops the Artificial Analysis Intelligence IndexARTICLEArtificial Analysis Claude Opus 5.5 (Adaptive Reasoning, High Effort, Default Fallback)EVALUATIONIntelligence Index 54 Claude Opus 5.5 (Adaptive Reasoning, Max Effort, Default Fallback)EVALUATIONIntelligence Index 58 Claude Opus 5.5 (Adaptive Reasoning, Medium Effort, Default Fallback)EVALUATIONIntelligence Index 51 Claude Opus 5.5 (Adaptive Reasoning, Xhigh Effort, Default Fallback)EVALUATIONIntelligence Index 56 Claude Opus 5.5 (High, Default Fallback)EVALUATIONIntelligence Index 54 Claude Opus 5.5 (Max, Default Fallback)EVALUATIONIntelligence Index 58 Claude Opus 5.5 (Medium, Default Fallback)EVALUATIONIntelligence Index 51 Claude Opus 5.5 (Xhigh, Default Fallback)EVALUATIONIntelligence Index 56 Claude Opus 5.5 takes the top spot on the Artificial Analysis Intelligence IndexARTICLEArtificial Analysis Claude Sonnet 5.5 (Adaptive Reasoning, High Effort, Default Fallback)EVALUATIONIntelligence Index 47 Claude Sonnet 5.5 (Adaptive Reasoning, Max Effort, Default Fallback)EVALUATIONIntelligence Index 56 Claude Sonnet 5.5 (Adaptive Reasoning, Xhigh Effort, Default Fallback)EVALUATIONIntelligence Index 52 Claude Sonnet 5.5 (High, Default Fallback)EVALUATIONIntelligence Index 47 Claude Sonnet 5.5 (Max, Default Fallback)EVALUATIONIntelligence Index 56 Claude Sonnet 5.5 (Xhigh, Default Fallback)EVALUATIONIntelligence Index 52 Claude Sonnet 5.5 reaches #2 on the Artificial Analysis Intelligence IndexARTICLEArtificial Analysis Gemini 4 Argon (High)EVALUATIONIntelligence Index 53 Gemini 4 Argon: Google is back as one of the top three labs in intelligence achievedARTICLEArtificial Analysis Google has released Gemini 3.8 Flash, its fourth Flash model in under four monthsARTICLEArtificial Analysis GPT-6 Astra (High)EVALUATIONIntelligence Index 51 GPT-6 Astra (Low)EVALUATIONIntelligence Index 46 GPT-6 Astra (Max)EVALUATIONIntelligence Index 53 GPT-6 Astra (Medium)EVALUATIONIntelligence Index 50 GPT-6 Astra (Xhigh)EVALUATIONIntelligence Index 52 GPT-6 Sol (Max)EVALUATIONIntelligence Index 48 GPT-6 Sol and Luna push the cost efficiency frontierARTICLEArtificial Analysis GPT-6.1 Sol (High)EVALUATIONIntelligence Index 50 GPT-6.1 Sol (Max)EVALUATIONIntelligence Index 52 GPT-6.1 Sol (Medium)EVALUATIONIntelligence Index 48 GPT-6.1 Sol (Xhigh)EVALUATIONIntelligence Index 51 GPT-6.1 Sol replaces GPT-6 Sol after just 7 days, with near-Astra intelligenceARTICLEArtificial Analysis Grok 4.7 (High)EVALUATIONIntelligence Index 46 Grok 4.7 (Xhigh)EVALUATIONIntelligence Index 46 KKorean AI Lab Upstage releases Solar Mini 4ARTICLEArtificial Analysis MiMo-V2.6-ProEVALUATIONIntelligence Index 46 Muse Spark 1.3 (Max)EVALUATIONIntelligence Index 48 Muse Spark 1.3 (Xhigh)EVALUATIONIntelligence Index 45 Muse Spark 1.3: Meta reaches the frontierARTICLEArtificial Analysis OOpenBMB releases MiniCPM5-2BARTICLEArtificial Analysis Qwen3.8 Max (0902)EVALUATIONIntelligence Index 45

27 Aug 2026

AAgnes AI releases Agnes 2.5 Pro BetaARTICLEArtificial Analysis Qwen3.8-Flash-NextEVALUATIONIntelligence Index 56

26 Aug 2026

AAgnes 2.5 Pro BetaEVALUATIONIntelligence Index 49 GLM-5.3-FlashEVALUATIONIntelligence Index 57