# AKI AI & Automation Intelligence Index™ (v1.1)

**Citation Object**: `aki:automation:top100` | **Edition**: 2026.4.0 | **Updated**: 2026-10-12

> Independent actuarial intelligence across autonomous agents, tools, workflows, frontier models, inference fabric, and research synthesis.

## Product & Agent Leaderboard (36 Verified Systems)

| Rank | Category | Product | Vendor | Release | Total Score | Capability | Quality | Reliability | Governance | Status |
| :--- | :--- | :--- | :--- | :--- | :--- | :--- | :--- | :--- | :--- | :---
| #1 | ai-tools | **Cursor IDE Agent** | Anysphere | `v0.45-prod` | **91.5** / 100 | 94.0 | 92.0 | 91.0 | 86.0 | TESTED |
| #2 | ai-tools | **Windsurf IDE** | Codeium | `v1.2-prod` | **90.7** / 100 | 93.0 | 91.0 | 90.0 | 85.0 | TESTED |
| #3 | ai-tools | **v0 Generative UI** | Vercel | `v2026.3` | **90.9** / 100 | 92.0 | 90.0 | 92.0 | 88.0 | VERIFIED |
| #4 | ai-tools | **Bolt.new Web Sandbox** | StackBlitz | `v2.1` | **89.6** / 100 | 91.0 | 89.0 | 89.0 | 84.0 | TESTED |
| #5 | ai-tools | **Replit Agent** | Replit | `v2026-Q3` | **87.8** / 100 | 90.0 | 88.0 | 87.0 | 83.0 | VERIFIED |
| #6 | ai-tools | **GitHub Copilot Workspace** | GitHub / Microsoft | `v2026.4` | **88.9** / 100 | 89.0 | 87.0 | 92.0 | 90.0 | DOCUMENTED |
| #1 | ai-agents | **Claude Code CLI Agent** | Anthropic | `v1.2-beta` | **92.8** / 100 | 96.0 | 95.0 | 92.0 | 90.0 | TESTED |
| #2 | ai-agents | **Devin Software Engineer** | Cognition AI | `v2.0-ent` | **87.2** / 100 | 91.0 | 89.0 | 86.0 | 88.0 | TESTED |
| #3 | ai-agents | **AutoGPT Forge** | Significant Gravitas | `v0.6-oss` | **87.1** / 100 | 88.0 | 85.0 | 82.0 | 86.0 | TESTED |
| #4 | ai-agents | **CrewAI Enterprise** | CrewAI Inc | `v0.75` | **89.8** / 100 | 90.0 | 88.0 | 89.0 | 89.0 | VERIFIED |
| #5 | ai-agents | **Semantic Kernel Agents** | Microsoft | `v1.32` | **90.3** / 100 | 89.0 | 88.0 | 94.0 | 95.0 | VERIFIED |
| #6 | ai-agents | **SWE-Agent Autonomous Bench** | Princeton NLP | `v1.1-oss` | **89.7** / 100 | 92.0 | 91.0 | 85.0 | 87.0 | TESTED |
| #1 | ai-automation | **Make AI Workflow Hub** | Celonis / Make | `v2026.3` | **90.3** / 100 | 88.0 | 86.0 | 94.0 | 92.0 | VERIFIED |
| #2 | ai-automation | **n8n AI Advanced Orchestration** | n8n GmbH | `v1.68` | **91.6** / 100 | 90.0 | 88.0 | 92.0 | 94.0 | TESTED |
| #3 | ai-automation | **Zapier Central Autonomous Bots** | Zapier | `v2026.2` | **89.8** / 100 | 87.0 | 85.0 | 93.0 | 91.0 | VERIFIED |
| #4 | ai-automation | **Activepieces Open Automation** | Activepieces | `v0.32` | **90.0** / 100 | 86.0 | 84.0 | 90.0 | 92.0 | TESTED |
| #5 | ai-automation | **Langflow Visual Framework** | DataStax / Langflow | `v1.1` | **90.7** / 100 | 89.0 | 87.0 | 88.0 | 90.0 | TESTED |
| #6 | ai-automation | **Flowise Low-Code Hub** | Flowise OSS | `v2.2` | **89.9** / 100 | 88.0 | 86.0 | 87.0 | 89.0 | TESTED |
| #1 | ai-models | **Claude 3.5 Sonnet** | Anthropic | `Oct-2024` | **92.5** / 100 | 95.0 | 94.0 | 93.0 | 92.0 | TESTED |
| #2 | ai-models | **GPT-4o Multimodal** | OpenAI | `May-2024` | **90.4** / 100 | 92.0 | 91.0 | 90.0 | 88.0 | TESTED |
| #3 | ai-models | **DeepSeek-V3 Reasoning Core** | DeepSeek | `Dec-2024` | **92.6** / 100 | 94.0 | 93.0 | 91.0 | 89.0 | TESTED |
| #4 | ai-models | **Gemini 1.5 Pro 2M** | Google DeepMind | `Sep-2024` | **91.0** / 100 | 93.0 | 92.0 | 91.0 | 91.0 | TESTED |
| #5 | ai-models | **Llama 3.1 405B Instruct** | Meta | `Jul-2024` | **90.9** / 100 | 92.0 | 90.0 | 92.0 | 91.0 | TESTED |
| #6 | ai-models | **Qwen 2.5 72B Instruct** | Alibaba Cloud | `Sep-2024` | **90.4** / 100 | 91.0 | 89.0 | 90.0 | 90.0 | TESTED |
| #1 | ai-infrastructure | **Langfuse LLM Observability** | Langfuse OSS | `v3.2` | **93.4** / 100 | 92.0 | 90.0 | 95.0 | 94.0 | TESTED |
| #2 | ai-infrastructure | **Arize Phoenix Open Tracing** | Arize AI | `v4.18` | **92.4** / 100 | 91.0 | 89.0 | 94.0 | 93.0 | TESTED |
| #3 | ai-infrastructure | **Braintrust Enterprise Eval** | Braintrust Data | `v2026.3` | **91.5** / 100 | 93.0 | 92.0 | 93.0 | 95.0 | VERIFIED |
| #4 | ai-infrastructure | **Weights & Biases Weave** | Weights & Biases | `v0.51` | **90.8** / 100 | 90.0 | 89.0 | 93.0 | 94.0 | VERIFIED |
| #5 | ai-infrastructure | **Promptfoo Security & Eval** | Promptfoo OSS | `v0.98` | **92.7** / 100 | 92.0 | 91.0 | 91.0 | 96.0 | TESTED |
| #6 | ai-infrastructure | **Helicone LLM Gateway & Cache** | Helicone | `v2.8` | **91.7** / 100 | 89.0 | 88.0 | 95.0 | 91.0 | TESTED |
| #1 | ai-research | **Perplexity Enterprise Pro** | Perplexity AI | `v2026-Q3` | **89.8** / 100 | 90.0 | 92.0 | 89.0 | 88.0 | VERIFIED |
| #2 | ai-research | **Elicit Research Assistant** | Elicit Inc | `v2026.2` | **90.1** / 100 | 91.0 | 93.0 | 88.0 | 89.0 | TESTED |
| #3 | ai-research | **Consensus Evidence Search** | Consensus AI | `v2026.1` | **89.7** / 100 | 89.0 | 92.0 | 89.0 | 88.0 | VERIFIED |
| #4 | ai-research | **Scite.ai Citation Context** | Research Solutions | `v2026.3` | **89.6** / 100 | 88.0 | 91.0 | 90.0 | 90.0 | VERIFIED |
| #5 | ai-research | **Semantic Scholar AI Reader** | Allen Institute for AI | `v2026-oss` | **89.9** / 100 | 87.0 | 90.0 | 88.0 | 92.0 | TESTED |
| #6 | ai-research | **Stanford STORM Research Agent** | Stanford OVAL | `v1.2-oss` | **89.3** / 100 | 91.0 | 90.0 | 85.0 | 89.0 | TESTED |

## Change Observatory (Forensic Event Ledger)

| Date | Product | Category | Event Type | Previous Value | New Value | Delta | Evidence |
| :--- | :--- | :--- | :--- | :--- | :--- | :--- | :---
| 2026-10-08 | **Cursor IDE Agent** | ai-tools | `capability_gain` | Agent mode single-file edit limit | **Full workspace multi-file synthesis** | +4.5 | [Evidence](https://aki1k.com/evidence/changes/cursor-v045.diff) |
| 2026-10-06 | **Windsurf IDE** | ai-tools | `capability_gain` | Standard file completion | **Cascade flow autonomous context buffer** | +3.8 | [Evidence](https://aki1k.com/evidence/changes/windsurf-cascade.diff) |
| 2026-10-04 | **v0 Generative UI** | ai-tools | `pricing_shift` | Credit pack paywall for Next.js exports | **Unlimited standard project exports on team tier** | +5.2 | [Evidence](https://aki1k.com/evidence/changes/v0-pricing-2026.html) |
| 2026-10-05 | **Claude Code CLI Agent** | ai-agents | `pricing_shift` | $0.03/1k input flat tier | **$0.003/1k input prompt caching tier** | +8.2 | [Evidence](https://aki1k.com/evidence/changes/claude-cache-announcement.html) |
| 2026-10-02 | **Devin Software Engineer** | ai-agents | `governance_update` | Internal cloud sandbox only | **Enterprise VPC isolated deployment** | +6.5 | [Evidence](https://aki1k.com/evidence/changes/devin-enterprise-vpc.pdf) |
| 2026-09-28 | **SWE-Agent Autonomous Bench** | ai-agents | `capability_gain` | Bash execution timeout at 30s | **Subprocess streaming timeout at 120s with crash recovery** | +4.0 | [Evidence](https://aki1k.com/evidence/changes/swe-agent-pr412.patch) |
| 2026-10-01 | **n8n AI Advanced Orchestration** | ai-automation | `governance_update` | Community unverified credential storage | **SOC2 Type II verified enterprise vault** | +6.0 | [Evidence](https://aki1k.com/evidence/changes/n8n-soc2.pdf) |
| 2026-09-25 | **Make AI Workflow Hub** | ai-automation | `regression_detected` | Zero data loss under burst webhook traffic | **2.1% dropped webhooks observed during regional spike** | -3.2 | [Evidence](https://aki1k.com/evidence/changes/make-incident-sep26.log) |
| 2026-09-20 | **Activepieces Open Automation** | ai-automation | `capability_gain` | Sequential node execution only | **Parallel branch execution with asynchronous promise joins** | +4.8 | [Evidence](https://aki1k.com/evidence/changes/activepieces-v032.diff) |
| 2026-10-02 | **Claude 3.5 Sonnet** | ai-models | `capability_gain` | June 2024 checkpoint baseline | **October 2024 refreshed checkpoint with Computer Use API** | +5.5 | [Evidence](https://aki1k.com/evidence/changes/sonnet-oct-release.pdf) |
| 2026-10-04 | **DeepSeek-V3 Reasoning Core** | ai-models | `pricing_shift` | $0.55/1M token API pricing | **$0.14/1M input and $0.28/1M output tokens** | +9.1 | [Evidence](https://aki1k.com/evidence/changes/deepseek-pricing-v3.html) |
| 2026-09-22 | **GPT-4o Multimodal** | ai-models | `regression_detected` | Strict tool-calling JSON schema adherence | **Edge-case schema violation observed in nested array objects** | -2.4 | [Evidence](https://aki1k.com/evidence/changes/gpt4o-json-regression.log) |
| 2026-09-30 | **Langfuse LLM Observability** | ai-infrastructure | `capability_gain` | Synchronous span batch flush | **Zero-blocking asynchronous ClickHouse ingestion** | +4.2 | [Evidence](https://aki1k.com/evidence/changes/langfuse-clickhouse.patch) |
| 2026-10-03 | **Promptfoo Security & Eval** | ai-infrastructure | `capability_gain` | Static test assertions | **LLM-assisted adaptive red teaming evaluation** | +5.0 | [Evidence](https://aki1k.com/evidence/changes/promptfoo-redteam.diff) |
| 2026-09-24 | **Helicone LLM Gateway & Cache** | ai-infrastructure | `pricing_shift` | Per-seat enterprise plan requirement | **Free open-source self-hosting license ratified** | +6.8 | [Evidence](https://aki1k.com/evidence/changes/helicone-oss-license.md) |
| 2026-10-07 | **Perplexity Enterprise Pro** | ai-research | `governance_update` | Standard web search logging | **Zero data retention enterprise agreement guarantee** | +5.5 | [Evidence](https://aki1k.com/evidence/changes/perplexity-enterprise-terms.pdf) |
| 2026-09-29 | **Elicit Research Assistant** | ai-research | `capability_gain` | 20-paper batch extraction limit | **1,000-paper asynchronous systematic synthesis** | +6.2 | [Evidence](https://aki1k.com/evidence/changes/elicit-batch1k.html) |
| 2026-09-26 | **Consensus Evidence Search** | ai-research | `capability_gain` | Keyword search relevance | **Academic semantic graph vector search with DOI verification** | +4.1 | [Evidence](https://aki1k.com/evidence/changes/consensus-doi.diff) |

## 9-Metric Actuarial Rubric Weights

| Metric | Weight | Scope & Operational Definition |
| :--- | :--- | :---
| **Capability & Task Execution** | 18% | Raw reasoning, multi-turn task resolution, and complex execution capability under stress testing. |
| **Output Quality & Fidelity** | 15% | Precision of delivered artifacts, syntactic correctness, and fidelity to structured instructions. |
| **Reliability & Error Recovery** | 15% | Uptime, state recovery after network partition or subprocess failure, and deterministic output repeatability. |
| **Developer Usability & DX** | 10% | Ergonomics of CLI or GUI interfaces, setup velocity, error messaging clarity, and documentation quality. |
| **Integration & Interoperability** | 12% | Breadth and stability of third-party API connectors, MCP protocol support, and containerization standards. |
| **Transparency & Observability** | 8% | Audit log granularity, OpenTelemetry export support, open-source weight inspection, and reproducible prompt tracing. |
| **Enterprise Governance & Security** | 10% | Data isolation, SOC2/ISO compliance, red team injection defense, and private VPC deployment support. |
| **Cost Efficiency & Token Economics** | 6% | Unit compute and token expenditure normalized per completed verified task. |
| **Recency & Maintenance Velocity** | 6% | Recency of verified releases, active repository maintenance, and empirical recency decay. |