WICK AI Agentic Tech Stack

State of the art for every slot in the stack. Auto-updated, AI-reviewed weekly.

CEO / Orchestrator

Top-level orchestration model for the Green Wick AI agent swarm. Ranked by agentic benchmarks (SWE-bench, GAIA, AgentBench, BFCL) — not just raw LLM reasoning. Framework scaffolding matters as much as the model.

Current: Claude Code (Opus 4.8) + native tool-use
#1 Claude Code (Opus 4.8) + native tool-use (98.5)
#2 Claude Code (Sonnet 4.6) + native tool-use (95.3)
#3 Codex CLI + GPT-5.5 (xhigh) (94.4)
2026-10-06
LLM General Reasoning

General-purpose reasoning model for the OpenClaw agent swarm

Current: nvidia/nemotron-3-super-120b-a12b:free
#1 google/gemma-4-31b-it:free (94.9) Free on OpenRouter · no retention
#2 google/gemma-4-26b-a4b-it:free (93.3) Free on OpenRouter · no retention
#3 inclusionai/ling-3.1-flash (80.3) Free on OpenRouter · no retention
2026-10-06

Local model for OpenClaw agent reasoning (ollama / llama.cpp)

Current: Qwen 3.5 27B
#1 QwQ-32B (69.5) Free from Huggingface · no retention
#2 Gemma-3-27B-it (61.7) Free from Huggingface · no retention
#3 Nemotron-3-Nano-30B-A3B (60.6) Free from Huggingface · no retention
2026-10-06
Voice Transcription (STT)

Local ASR model for voice note transcription (faster-whisper etc.)

Current: faster-whisper (base.en)
#1 ggerganov/whisper.cpp (78.7)
#2 SYSTRAN/faster-whisper (74.3)
#3 openai/whisper (69.5)
2026-10-06
Meta Search

Self-hosted search aggregator (currently SearxNG)

Current: SearxNG
#1 searxng/searxng (82.4)
#2 ItzCrazyKns/Perplexica (78.2)
#3 benbusby/whoogle-search (37.7)
2026-10-06
Browser Automation

Browser control for AI agents (Playwright, browser-use, etc.)

Current: Playwright
#1 browser-use/browser-use (96.4)
#2 microsoft/playwright (72.3)
#3 browserbase/stagehand (65.0)
2026-10-06
Image Generation

Free-tier API for AI image generation (Flux, DALL-E, Stable Diffusion etc.)

Current: Flux.1 schnell (Replicate)
#1 Flux.1 [schnell] (Replicate free) (100.0)
#2 Stable Diffusion 3.5 (Stability free) (90.0)
#3 DALL-E 3 (Bing Image Creator) (80.0)
2026-10-06

Local diffusion model for image generation on consumer GPU (ComfyUI, Forge etc.)

Current: ComfyUI + SDXL
#1 comfyanonymous/ComfyUI (84.7)
#2 AUTOMATIC1111/stable-diffusion-webui (50.0)
#3 lllyasviel/stable-diffusion-webui-forge (3.4)
2026-10-06
Email Verification

Email verification service with best free tier quota and accuracy

Current: EmailListVerify
#1 QuickEmailVerification (100.0)
#2 EmailListVerify (85.0)
#3 Abstract API (70.0)
2026-10-06
Agentic Orchestration

Multi-agent orchestration frameworks for building AI agent systems (LangGraph, CrewAI, etc.)

Current: LangGraph
#1 n8n-io/n8n (91.1)
#2 langchain-ai/langgraph (74.5)
#3 crewAIInc/crewAI (73.3)
2026-10-06

Specific Use Cases

LLM Code Review

Code analysis model for the Vibe Audit SaaS product

Current: qwen/qwen3-coder:free
#1 apodex/apodex-1.1-mini:free (99.7) Free on OpenRouter · no retention
#2 thinkingmachines/inkling-small:free (93.9) Free on OpenRouter · no retention
#3 thinkingmachines/inkling:free (93.9) Free on OpenRouter · no retention
2026-10-06
LLM Chat Analysis

Summarization and extraction model for Telegram chat analysis

Current: qwen/qwen3-coder:free
#1 inclusionai/ling-3.1-flash (84.8) Free on OpenRouter · no retention
#2 nvidia/nemotron-3-ultra-550b-a55b:free (76.6) Free on OpenRouter · no retention
#3 apodex/apodex-1.1-mini:free (69.2) Free on OpenRouter · no retention
2026-10-06

Data Source Health

Source Status Checked
OpenRouter API UP (200) 2026-10-07
HuggingFace Hub UP (200) 2026-10-07
Arena.ai (Chatbot Arena) UP (200) 2026-10-07
Artificial Analysis UP (200) 2026-10-07
Vals.ai UP (200) 2026-10-07
ZeroEval UP (200) 2026-10-07