API providerUsage billing; see current official prices
DeepSeek API
Official model API; the documentation lists deepseek-flash and deepseek-v4-pro.
AQI AI / SEARCH
Search all resources
Search recorded models, integrations, learning paths, workflows and news. Filtering runs in your browser.
Filter research
154 results
HarnessFree software; check license and compute costs
DeepSeek Harness
An official open-source agent harness from DeepSeek AI with a plugin architecture, in developer preview.
AgentFree software; check license and compute costs
Codex CLI
OpenAI’s official local command-line coding agent.
AgentNot verified
Claude Code
Anthropic’s coding agent for terminal and development workflows.
AgentFree software; check license and compute costs
Cline
A coding agent supporting multiple providers, Ollama, LM Studio and MCP.
AgentFree software; check license and compute costs
Continue
An open-source coding agent project for development workflows.
AgentFree software; check license and compute costs
Aider
Terminal AI pair programming with documented cloud and local model connections.
AgentFree software; check license and compute costs
Roo Code
An open-source AI agent project for the code editor.
AgentNot verified
Cursor
An AI editor with documentation for agents, rules, MCP and development workflows.
AgentFree software; check license and compute costs
OpenClaw
A self-hosted personal AI assistant project with platform and channel integrations.
Local runtimeFree software; check license and compute costs
Ollama
A local model runtime and API tool supporting a subset of OpenAI endpoints.
Local runtimeFree software; check license and compute costs
LM Studio
A desktop local-model tool with compatible Chat Completions and Responses endpoints.
Local runtimeFree software; check license and compute costs
vLLM
A model-serving inference engine with an OpenAI-compatible HTTP server.
Local runtimeFree software; check license and compute costs
llama.cpp
A C/C++ language-model inference project for multiple local hardware backends.
ModelFree software; check license and compute costs
Qwen3
An open model family released by Alibaba’s Qwen team.
API providerUsage billing; see current official prices
Gemini API
Official Google Gemini API; gemini-4-argon became generally available on 2026-10-13, with time-limited introductory pricing.
ModelNot verified
Llama models
Meta’s official repository of Llama model materials and utilities.
ModelNot verified
Mistral models
Mistral’s official model catalog covers different uses and deployment conditions.
Device toolFree software; check license and compute costs
Android Debug Bridge
Android’s official debugging and device communication tool, requiring device authorization.
Integration platformFree software; check license and compute costs
scrcpy
An open-source tool to mirror and control an authorized Android device from a computer.
Integration platformFree software; check license and compute costs
Playwright
Browser automation and end-to-end testing for interactions on controlled pages.
Integration platformNot verified
Telegram Bot API
Telegram’s official HTTP API for bots.
Integration platformNot verified
Discord Developer Platform
Discord’s official platform for apps, bots and interactions.
Integration platformNot verified
Slack API
Slack’s official app and automation developer documentation.
Integration platformFree software; check license and compute costs
Home Assistant
A home automation platform with an official Ollama conversation integration.
Integration platformFree software; check license and compute costs
OBS WebSocket
The official project for controlling OBS Studio through WebSocket.
Integration platformNot verified
n8n
An integration platform with self-hosting and workflow automation documentation.
Integration platformFree software; check license and compute costs
ComfyUI
An open-source interface and backend organizing generative workflows through node graphs.
Integration platformFree software; check license and compute costs
Unity ML-Agents
Unity’s official open-source ML toolkit for training and testing agents in games and simulations.
Integration platformFree software; check license and compute costs
Godot HTTPRequest
Godot’s official HTTPRequest interface connects external services; AI integration needs custom implementation.
Integration platformFree software; check license and compute costs
FFmpeg
Audio/video tools and libraries with official documentation for codecs, formats and filters.
ModelFree software; check license and compute costs
Whisper
OpenAI’s open-source speech recognition models and transcription tool.
ProtocolFree software; check license and compute costs
Model Context Protocol
An open protocol connecting AI applications with external tools and data.
Integration platformNot verified
Microsoft Graph
The official Microsoft 365 data API covering mail, calendars, Excel, OneDrive and other services.
Integration platformNot verified
LINE Messaging API
The official bot interface using LINE Official Accounts and webhooks.
Integration platformNot verified
QQ Open Platform
Tencent’s official QQ platform with bot registration and development materials.
Integration platformNot verified
Apple Shortcuts
Apple’s official multi-step action automation for supported iPhone and iPad apps and services.
Device toolNot verified
Raspberry Pi
Official board documentation for camera, sensor and edge-project research.
API providerUsage billing; see current official prices
OpenAI API
Official OpenAI model API documenting gpt-6.1-sol; prices vary by model, context and service tier.
API providerUsage billing; see current official prices
Anthropic Claude API
Official Anthropic Claude API documenting Claude Sonnet 5.5 and Opus 5.5.
API providerUsage billing; see current official prices
Alibaba Cloud Model Studio
Alibaba Cloud hosted model platform; Qwen API prices, free quotas and deployment scopes vary by region.
API providerUsage billing; see current official prices
Z.ai GLM API
Official Z.ai API; pricing lists GLM-5.3 and GLM-5.3-Flash, while exact request IDs remain unverified.
API providerUsage billing; see current official prices
Kimi API
Official Kimi API documenting kimi-k3, vision input and a 1M-token context; billed separately from memberships and Kimi Code.
API providerUsage billing; see current official prices
MiniMax API
Official MiniMax API for language, audio, image and video services, each with distinct billing units.
GatewayUsage billing; see current official prices
OpenRouter
Multi-provider model router; verify model provenance, logging and billing for the selected route.
ModelNot verified
Gemma
Google open model family including documented Gemma 4; check each artifact, license and hardware result separately.
Integration platformNot verified
ROS 2 / Gazebo ros_gz
Read sensors in isolated simulation before evaluating AI planning; leave real actuators disconnected.
MCPServerNot verified
Slack MCP Server
Official Slack MCP uses Streamable HTTP and OAuth; only published or internal apps qualify. Check scopes and rate limits per tool.
PluginNot verified
Slack MCP and Skills Plugin
Official Slack plugin bundles development Skills; its Codex install provides Skills only, without the Slack MCP workspace connection or slack-search.
SkillNot verified
slack-api Skill
slack-api is an API guidance Skill in the official plugin, covering methods, scopes, pagination and errors; it does not itself grant account access.
API providerNot verified
BigModel API (China)
China BigModel documentation provides glm-5 request examples and lists a 200K context window and 128K maximum output. This site did not call the API.
PluginNot verified
dsh-bili-scroll
Community video-search, subtitle and report plugin, with author video/source links; not an official DeepSeek plugin.
Community reference · not a default install recommendationPluginNot verified
dsh-wechat-plugin
Community WeChat iLink channel dsh-wechat-plugin; distinct from public accounts/WeCom, not an official DSH plugin.
Community reference · not a default install recommendationPluginNot verified
@sugarforever/dsh-lark
Community plugin using Feishu/Lark SDK WebSocket for messages and replies; custom app/approval required, not platform-official.
Community reference · not a default install recommendationMobile & devicesNot verified
Observe and control Android from a PC
Establish a revocable authorized connection and mirroring before investigating an AI decision layer.
Game developmentNot verified
Design local NPC dialogue for your Godot game
Send test dialogue to a local model over HTTP; connection code is custom work.
Chat platformsNot verified
Design a local-model bot with the official Bot API
A bot program bridges Telegram and a local model; both APIs were reviewed, but the bridge was not tested.
Smart homeNot verified
Start a local home assistant with one light
Use the official Home Assistant Ollama integration, starting with conversation and then limited control.
BrowserNot verified
Use browser automation to test your own pages
Use Playwright locators, interactions and assertions on owned sites without presuming authority for external actions.
Office softwareNot verified
Start office integration with read-only calendar summaries
Access cloud office data through Microsoft Graph permissions; models process authorized fields only.
Creative toolsNot verified
Scope OBS control for a live assistant
Verify OBS WebSocket authentication and a test scene first; develop the AI bridge separately.
Mobile & devicesNot verified
Design iOS automation with Apple actions
Compose actions permitted by the system and apps with Shortcuts; this does not provide unrestricted phone remote control.
Chat platformsNot verified
Discord official command bot
Bound inputs with slash commands before adding a separate model layer.
Chat platformsNot verified
Slack Socket Mode local bot
Receive authorized events over an outbound WebSocket; test fixed replies first.
Chat platformsNot verified
QQ official bot integration
Receive and reply to authorized messages using an official bot identity.
HardwareNot verified
ROS 2 and Gazebo simulation observation
Read sensors in isolated simulation before evaluating AI planning; leave real actuators disconnected.
Office softwareNot verified
Slack Skills and MCP integration boundaries
Separate development instruction packages from authorized workspace connections, then verify each client.
Chat platformsNot verified
WeCom group notifications: Webhook quick guide
Send AI summaries to a designated group; start with text, then consider rich content and files.
Chat platformsNot verified
WeChat AI integration: choose your route
WeCom AI bots, OpenClaw and public accounts are separate routes; author examples shorten the documentation search.
Chat platformsNot verified
WeCom capability map: mini programs, web and office
Choose by runtime; distinguish screenshot navigation, author examples and unverified API-specific permissions.
Chat platformsNot verified
DSH and Bilibili: start with a manual report
Community video-search, subtitle and report plugin, with author video/source links; not an official DeepSeek plugin.
Chat platformsNot verified
Connect WeChat to DSH: permissions first
Community WeChat iLink channel dsh-wechat-plugin; distinct from public accounts/WeCom, not an official DSH plugin.
Chat platformsNot verified
DSH and Feishu/Lark: bot and allowlists
Community plugin using Feishu/Lark SDK WebSocket for messages and replies; custom app/approval required, not platform-official.
Research guideNot verified
From AI chat to a local endpoint
Learn the roles of models, runtimes, APIs and agents, then complete a local request.
Research guideNot verified
Getting started with local coding agents
Verify the model, then practice permissions and review on reversible code changes.
Research guideNot verified
AI video basics: transcription and subtitles
Begin with owned media and practice subtitle review, synchronization and rights records.
Research guideNot verified
From authorized Android connection to AI research
Learn connection, state checks and mirroring before scoped AI control.
Research guideNot verified
AI prototypes in your own game
Distinguish NPC dialogue from reinforcement learning and build a dialogue scene with fallback.
Research guideNot verified
Official chatbots and privacy boundaries
Understand bot APIs versus personal accounts and start with drafts for limited test users.
Research guideNot verified
Minimum permissions for a local home assistant
Progress from conversation to one light while keeping manual control and a shutdown path.
Research guideNot verified
AI images: a traceable experiment
Learn workflows and model files, then compare prompts one variable at a time.
Research guideNot verified
AI video and short drama: finish three shots first
Connect original storyboards, short image-to-video tests, narration and editing through shot-continuity checks.
Research guideNot verified
AI narration: from original lines to an authorized track
Practice with preset or original designed voices, treating revocable voice permission separately; real-person cloning is not the default.
Research guideNot verified
AI music: separate code and model licenses first
Learn MusicGen conditioning and listening review in a noncommercial exercise; use separately cleared music for commercial projects.
Research guideNot verified
AI websites, Skills and read-only MCP basics
Separate content, reusable instructions and tool protocols in a small local site, then verify minimal permissions using synthetic data.
Research guideNot verified
Set up your first local chat endpoint
Verify local inference and its API before connecting an agent.
Research guideNot verified
Make a small code change with Cline and a local model
Start with an isolated repository and reversible changes, checking tool-call capability.
Research guideNot verified
Connect DeepSeek Harness through Ollama’s official launcher
Use an installed local model and distinguish inference from cloud-dependent web search.
Research guideNot verified
Observe and control Android from a PC
Establish a revocable authorized connection and mirroring before investigating an AI decision layer.
Research guideNot verified
Design local NPC dialogue for your Godot game
Send test dialogue to a local model over HTTP; connection code is custom work.
Research guideNot verified
Design a local-model bot with the official Bot API
A bot program bridges Telegram and a local model; both APIs were reviewed, but the bridge was not tested.
Research guideNot verified
Start a local home assistant with one light
Use the official Home Assistant Ollama integration, starting with conversation and then limited control.
Research guideNot verified
Create reviewable subtitles for your own short video
Transcribe with Whisper, review manually and process with FFmpeg; this does not claim to generate a new video.
Research guideNot verified
Run an image A/B test with an official workflow
Learn workflows and model files, then compare prompts one variable at a time.
Research guideNot verified
Create and review a three-shot short drama
Connect original storyboards, short image-to-video tests, narration and editing through shot-continuity checks.
Research guideNot verified
Local endpoint test: Ollama and Qwen3 8B
Two endpoints on an existing Windows installation passed synthetic checks; the first empty compatible response is retained as a failure.
Documentation check
DeepSeek Harness official provenance confirmed
The official organization repository and DeepSeek API docs cross-link it; the product remains a developer preview.
Documentation check
Ollama documents a Harness launcher
Models can be selected, but web search requires cloud access; documented support is not a site test.
Documentation check
Codex local-provider boundaries recorded
Official advanced config includes --oss with Ollama and LM Studio; model capabilities still need individual verification.
Documentation check
Check vLLM authentication coverage separately
Official docs warn API keys protect only certain paths; exposed services need additional isolation and access control.
Official releases
GPT-Live adds native Decisions integration
Applications can delegate natively from GPT-Live or trigger Decisions from transcript deltas; no API runtime test by this site.
Official releases
Claude Opus 5.5 fast mode is generally available
Opus 5.5 fast mode no longer requires preview access or a beta header; fast mode for Opus 5 and 4.8 remains preview.
Official releases
Claude launches OpenTelemetry export beta
Messages API requests and Managed Agents session traces can be exported to observability backends; currently beta.
Official releases
Gemini 4 Argon is generally available
Official Google Gemini API; gemini-4-argon became generally available on 2026-10-13, with time-limited introductory pricing.
Official releases
Ollama 0.40.5 fixes custom-tool conversation continuation
Official notes document a custom-tool continuation fix for compatible Responses and updated Nemotron support. Local endpoint checks used existing 0.34.4 and do not test this release.
Official releases
DeepSeek Harness updates alpha.4 prerelease
Official prerelease adds custom-model reasoning controls and custom plugin indexes. It remains alpha; review each plugin source and permission. Not run by this site.
Official releases
GPT-Live-1.1 rollout adds multilingual voice choices
Official October 15 notes confirm GPT-Live-1.1 and Quip are rolling out in ChatGPT Voice, with more language voices; availability varies by region and account.
Official releases
Z.ai separates Standard API and Coding Plan keys
From 2026-10-16, new Standard API and Coding Plan keys are billed by key type; changing endpoints does not change that type. Standard keys may charge the account balance rather than plan quota. Existing Compatible keys retain endpoint- and plan-history-dependent rules; this update does not require replacing them.
ModelFree software; check license and compute costs
bge-reranker-v2-m3
Embedding models produce retrieval vectors, rerankers score candidate passages, and speech models transcribe audio. These tasks are not comparable by chat quality. Batch, activation and audio memory are not modeled; specialist capacity stays unverified instead of using the chat-cache formula.
ModelFree software; check license and compute costs
DeepSeek-R1-Distill-Qwen-1.5B
Capacity estimate only; no automatic download or connection to your computer.
ModelFree software; check license and compute costs
DeepSeek-R1-Distill-Qwen-7B
Capacity estimate only; no automatic download or connection to your computer.
ModelFree software; check license and compute costs
DeepSeek-R1-Distill-Llama-8B
Capacity estimate only; no automatic download or connection to your computer.
ModelFree software; check license and compute costs
DeepSeek-R1-Distill-Qwen-14B
Capacity estimate only; no automatic download or connection to your computer.
ModelFree software; check license and compute costs
DeepSeek-R1-Distill-Qwen-32B
Capacity estimate only; no automatic download or connection to your computer.
ModelFree software; check license and compute costs
DeepSeek-R1-Distill-Llama-70B
Capacity estimate only; no automatic download or connection to your computer.
ModelFree software; check license and compute costs
gemma-3-1b-it
Capacity estimate only; no automatic download or connection to your computer.
ModelFree software; check license and compute costs
gemma-3-4b-it
Capacity estimate only; no automatic download or connection to your computer.
ModelFree software; check license and compute costs
gemma-3-12b-it
Capacity estimate only; no automatic download or connection to your computer.
ModelFree software; check license and compute costs
gemma-3-27b-it
Capacity estimate only; no automatic download or connection to your computer.
ModelFree software; check license and compute costs
gemma-4-12B
Capacity estimate only; no automatic download or connection to your computer.
ModelFree software; check license and compute costs
gpt-oss-20b
Capacity estimate only; no automatic download or connection to your computer.
ModelFree software; check license and compute costs
gpt-oss-120b
Capacity estimate only; no automatic download or connection to your computer.
ModelFree software; check license and compute costs
Llama-3.2-1B-Instruct
Capacity estimate only; no automatic download or connection to your computer.
ModelFree software; check license and compute costs
Llama-3.2-3B-Instruct
Capacity estimate only; no automatic download or connection to your computer.
ModelFree software; check license and compute costs
Llama-3.1-8B-Instruct
Capacity estimate only; no automatic download or connection to your computer.
ModelFree software; check license and compute costs
Llama-3.3-70B-Instruct
Capacity estimate only; no automatic download or connection to your computer.
ModelFree software; check license and compute costs
Ministral-3-3B-Instruct-2512-BF16
Capacity estimate only; no automatic download or connection to your computer.
ModelFree software; check license and compute costs
Ministral-3-8B-Instruct-2512-BF16
Capacity estimate only; no automatic download or connection to your computer.
ModelFree software; check license and compute costs
Ministral-3-14B-Instruct-2512-BF16
Capacity estimate only; no automatic download or connection to your computer.
ModelFree software; check license and compute costs
Mistral-Small-3.1-24B-Instruct-2503
Capacity estimate only; no automatic download or connection to your computer.
ModelFree software; check license and compute costs
Phi-4-mini-instruct
Capacity estimate only; no automatic download or connection to your computer.
ModelFree software; check license and compute costs
phi-4
Capacity estimate only; no automatic download or connection to your computer.
ModelFree software; check license and compute costs
Qwen3-0.6B
Capacity estimate only; no automatic download or connection to your computer.
ModelFree software; check license and compute costs
Qwen3-1.7B
Capacity estimate only; no automatic download or connection to your computer.
ModelFree software; check license and compute costs
Qwen3-4B
Capacity estimate only; no automatic download or connection to your computer.
ModelFree software; check license and compute costs
Qwen3-8B
Capacity estimate only; no automatic download or connection to your computer.
ModelFree software; check license and compute costs
Qwen3-14B
Capacity estimate only; no automatic download or connection to your computer.
ModelFree software; check license and compute costs
Qwen3-30B-A3B
Capacity estimate only; no automatic download or connection to your computer.
ModelFree software; check license and compute costs
Qwen3-32B
Capacity estimate only; no automatic download or connection to your computer.
ModelFree software; check license and compute costs
Qwen3-Coder-30B-A3B-Instruct
Capacity estimate only; no automatic download or connection to your computer.
ModelFree software; check license and compute costs
Qwen3-Embedding-0.6B
Embedding models produce retrieval vectors, rerankers score candidate passages, and speech models transcribe audio. These tasks are not comparable by chat quality. Batch, activation and audio memory are not modeled; specialist capacity stays unverified instead of using the chat-cache formula.
ModelFree software; check license and compute costs
Qwen3-Embedding-4B
Embedding models produce retrieval vectors, rerankers score candidate passages, and speech models transcribe audio. These tasks are not comparable by chat quality. Batch, activation and audio memory are not modeled; specialist capacity stays unverified instead of using the chat-cache formula.
ModelFree software; check license and compute costs
Qwen3-Embedding-8B
Embedding models produce retrieval vectors, rerankers score candidate passages, and speech models transcribe audio. These tasks are not comparable by chat quality. Batch, activation and audio memory are not modeled; specialist capacity stays unverified instead of using the chat-cache formula.
ModelFree software; check license and compute costs
Qwen3-Reranker-0.6B
Embedding models produce retrieval vectors, rerankers score candidate passages, and speech models transcribe audio. These tasks are not comparable by chat quality. Batch, activation and audio memory are not modeled; specialist capacity stays unverified instead of using the chat-cache formula.
ModelFree software; check license and compute costs
Qwen3-Reranker-4B
Embedding models produce retrieval vectors, rerankers score candidate passages, and speech models transcribe audio. These tasks are not comparable by chat quality. Batch, activation and audio memory are not modeled; specialist capacity stays unverified instead of using the chat-cache formula.
ModelFree software; check license and compute costs
Qwen3-Reranker-8B
Embedding models produce retrieval vectors, rerankers score candidate passages, and speech models transcribe audio. These tasks are not comparable by chat quality. Batch, activation and audio memory are not modeled; specialist capacity stays unverified instead of using the chat-cache formula.
ModelFree software; check license and compute costs
Qwen3.5-4B
Capacity estimate only; no automatic download or connection to your computer.
ModelFree software; check license and compute costs
Qwen3.5-9B
Capacity estimate only; no automatic download or connection to your computer.
ModelFree software; check license and compute costs
Qwen3.5-27B
Capacity estimate only; no automatic download or connection to your computer.
ModelFree software; check license and compute costs
Qwen3.8-27B
Capacity estimate only; no automatic download or connection to your computer.
ModelFree software; check license and compute costs
SmolLM2-135M-Instruct
Capacity estimate only; no automatic download or connection to your computer.
ModelFree software; check license and compute costs
SmolLM2-360M-Instruct
Capacity estimate only; no automatic download or connection to your computer.
ModelFree software; check license and compute costs
SmolLM2-1.7B-Instruct
Capacity estimate only; no automatic download or connection to your computer.
ModelFree software; check license and compute costs
whisper-large-v3-turbo
Embedding models produce retrieval vectors, rerankers score candidate passages, and speech models transcribe audio. These tasks are not comparable by chat quality. Batch, activation and audio memory are not modeled; specialist capacity stays unverified instead of using the chat-cache formula.
No matching records. Try clearing your filters.