As of 2026-10-08
Artificial Analysis.
Their board. Their ranks.
Artificial Analysis Intelligence, Agentic, and Coding indices, plus cost per task and speed.
AA #1 is Opus 5.5
Artificial Analysis
Copied from the AA current-status board (v4.3.2) and the Grok 4.6 / Opus 5.5 AA articles. No averaging.
| 1 | Opus 5.5 Anthropic · closed Anthropic’s 5.5 frontier flagship | 58Lab | — | — | $5.98Lab | — |
| 2 | Fable 5.1 Anthropic · closed New Claude 5 frontier tier | 53Lab | 58AA Agentic | 81.6AA Coding | $7.63Lab | — |
| 2 | GPT-6 Astra OpenAI · closed OpenAI GPT-6 flagship | 53Lab | 51.5AA Agentic | 76.9AA Coding | $3.26Lab | — |
| 4 | Opus 5 Anthropic · closed Anthropic’s flagship frontier model | 51Lab | 56.2AA Agentic | 78AA Coding | $5.86Lab | 53Lab |
| 5 | Fable 5 Anthropic · closed New Claude 5 frontier tier | 50Lab | 51AA Agentic | 76.5AA Coding | $8.75Lab | 67Lab |
| 6 | GPT-6 Sol OpenAI · closed GPT-6 flagship tier | 48Lab | — | — | — | — |
| 6 | Spark 1.3 Meta · closed Agent-native Meta model; reaches the frontier | 48Lab | 55.7AA Agentic | 75.8AA Coding | — | 222Lab |
| 8 | GPT-5.6 Sol OpenAI · closed Top-tier flagship reasoning | 47Lab | 50.5AA Agentic | 77.4AA Coding | $1.99Lab | 73Lab |
| 9 | Grok 4.7 SpaceXAI · closed Latest Grok frontier model | 46Lab | — | — | — | — |
| 9 | MiMo V2.6 Pro Xiaomi · open Open MIT; #1 open-weights on AA | 46Lab | — | — | — | — |
| 11 | Qwen3.8 Max Alibaba · closed Hosted 2.4T flagship; weights promised | 45Lab | 49.6AA Agentic | 71.8AA Coding | $5.41Lab | 39Lab |
| 11 | GLM-5.3 Z.ai · staged Latest GLM coding / cyber-defense | 45Lab | 53.4AA Agentic | 74.8AA Coding | $2.01Lab | 61Lab |
| 13 | Grok 4.6 SpaceXAI · closed Latest Grok frontier model | 44Lab | 53.4AA Agentic | 76.8AA Coding | $1.86Lab | 61Lab |
| 13 | Kimi K3 Moonshot AI · open Frontier open-weight | 44Lab | 50.6AA Agentic | 76.2AA Coding | $2.00Lab | 37Lab |
| 15 | GPT-5.6 Terra OpenAI · closed Balanced GPT-5.6 | 42Lab | 43.7AA Agentic | 76.7AA Coding | $1.40Lab | 83Lab |
| 15 | GLM-5.3 Flash Z.ai · open Open-weight GLM, MIT | 42Lab | — | — | $0.25Lab | — |
| 15 | Opus 4.8 Anthropic · closed Frontier reasoning/coding | 42Lab | 42.6AA Agentic | 74.3AA Coding | $4.08Lab | 57Lab |
| 18 | 3.8 Flash Google DeepMind · closed Latest Flash workhorse | 41Lab | 41.1AA Agentic | 76.3AA Coding | $1.24Lab | 283Lab |
| 18 | Opus 4.7 Anthropic · closed Frontier coding/reasoning | 41Lab | 39.5AA Agentic | 73.6AA Coding | — | — |
| 18 | Inkling Thinking Machines · open New open lab, Apache-2.0 | 41Lab | 24.3AA Agentic | 52.1AA Coding | $0.34Lab | 73Lab |
| 21 | Qwen3.8 open Alibaba · open Open-weight Qwen3.8 (AA #2 open) | 40Lab | — | — | $2.16Lab | 38Lab |
| 21 | Flash Next Alibaba · open Qwen4-architecture open preview | 40Lab | — | 73AA Coding | — | — |
| 21 | Spark 1.2 Meta · closed Agent-native Meta model | 40Lab | 44AA Agentic | 72.2AA Coding | $0.97Lab | 187Lab |
| 24 | 3.7 Flash Google DeepMind · closed Latest Flash workhorse | 39Lab | 36.4AA Agentic | 76.1AA Coding | $0.93Lab | 293Lab |
| 24 | Grok 4.5 SpaceXAI · closed Frontier coding/agents | 39Lab | 42.1AA Agentic | 72.5AA Coding | $1.04Lab | 57Lab |
| 24 | V4.1 Flash DeepSeek · open Open MoE successor to V4 Flash/Pro | 39Lab | — | — | $0.27Lab | 207Lab |
| 27 | Sonnet 5 Anthropic · closed Frontier general-purpose/agentic | 38Lab | 44.3AA Agentic | 71.5AA Coding | $5.09Lab | 79Lab |
| 28 | GPT-6 Luna OpenAI · closed Cost-efficient GPT-6 | 37Lab | — | — | — | — |
| 28 | GPT-5.6 Luna OpenAI · closed Cost-efficient GPT-5.6 | 37Lab | 42.7AA Agentic | 71.5AA Coding | $0.18Lab | 140Lab |
| 30 | V4 Pro 0813 DeepSeek · open Current DeepSeek-V4 open challenger | 36Lab | 49.6AA Agentic | 68.8AA Coding | $0.67Lab | 66Lab |
| 31 | V4 Flash DeepSeek · open Open price-performance pick | 35Lab | 41.7AA Agentic | 69.1AA Coding | $0.31Lab | 218Lab |
| 32 | GLM-5.2 Zhipu AI · open Frontier/open-weight coding | 34Lab | 39.4AA Agentic | 68.8AA Coding | $0.96Lab | 70Lab |
| 32 | 3.6 Flash Google DeepMind · closed Frontier-class fast reasoning | 34Lab | 30.1AA Agentic | 69.2AA Coding | $0.93Lab | 184Lab |
| 34 | 3.1 Pro Google DeepMind · closed Frontier multimodal/reasoning | 30Lab | 10.3AA Agentic | 68.8AA Coding | $0.67Lab | 118Lab |
| 35 | MiniMax M3 MiniMax · open Open-weight frontier-adjacent | 29Lab | 30.8AA Agentic | 58.6AA Coding | $0.51Lab | 106Lab |
| 35 | Qwen3.7 Max Alibaba · closed Frontier Chinese model | 29Lab | 23.9AA Agentic | 66AA Coding | $1.15Lab | 203Lab |
| 37 | GPT-5.5 OpenAI · closed Frontier reasoning + agents | 26Lab | 37.3AA Agentic | 74.9AA Coding | $0.69Lab | 127Lab |
| 37 | Grok 4.20 SpaceXAI · closed Frontier reasoning/agents | 26Lab | — | — | — | — |
| 37 | MiMo V2.5 Pro Xiaomi · open Open MIT, cheap per-task | 26Lab | 22.7AA Agentic | 60.2AA Coding | $0.03Lab | 46Lab |
| 37 | Hy3 Tencent · open Open Apache-2.0 coding/reasoning | 26Lab | 25.6AA Agentic | 58.8AA Coding | $0.04Lab | 64Lab |
| 41 | Grok 4.3 SpaceXAI · closed Frontier reasoning/agents | 25Lab | 17.2AA Agentic | 42.3AA Coding | — | — |
| 42 | LongCat 2.0 Meituan · open Frontier open-weight agentic coding | 19Lab | — | — | $0.06Lab | — |
| 43 | Doubao Seed ByteDance · closed Frontier reasoning | 17Prelim | — | — | — | — |
43 models on this board · # is this page’s rank