基準日 2026-07-22

AIランキング研究室

基準日 2026-07-22 · 21テーマ · Top20 · GitHub · Skills · X · Reddit · App Store · Play · リーダー · 研究所 · 論文 · PH · リンク索引
言語
品質ゲート 合格 幻覚防止: catalog map
テーマ — 基準日 2026-07-22
Cover for LLM基盤(2026-07)
カバー · Grok Imagine

LLM基盤(2026-07)

現行のフロンティア基盤モデル · top 20 · as_of 2026-07-22

情報源 / 公開API
#名称地図に載る理由
1Claude Fable 5 / Opus 4.8 class
マップ優先度: 100
Primary long-horizon control-plane class; still top-tier on agent/SWE boards
2GPT-5.6-Sol / GPT-5.5 Pro class
マップ優先度: 99
Coding/STEM flagship; default executor in Codex and ChatGPT product surface
3Gemini 3.1 / 3.x class
マップ優先度: 98
Long-context science/multimodal surface tightly tied to Google stack
4Grok 4.5 / xAI
マップ優先度: 97
Realtime + X-native retrieval; cost-effective cross-check route
5DeepSeek-V3 / R1 class
マップ優先度: 96
Open-weight reasoning cost leader for private/local deploy
6Qwen3 / Qwen 3.7 class (Alibaba)
マップ優先度: 95
Active bilingual + contest-math line; open and cloud dual track
7Llama 4 class (Meta)
マップ優先度: 94
Default open ecosystem base; wide fine-tune and private deploy
8Mistral Large / Mixtral line
マップ優先度: 93
EU-positioned open/API line favored in enterprise procurement
9Kimi K3 / Moonshot long-context
マップ優先度: 92
Long-context + strong frontend-code arena presence; CN document workflows
10GLM-5.x / Zhipu
マップ優先度: 91
CN API + open dual track; frequent open-weight Arena contender
11Amazon Nova class
マップ優先度: 90
Bedrock-native tiered models default inside AWS accounts
12Claude Sonnet 5 / writing tier
マップ優先度: 89
Writing/instruction-following product default; cost-efficient daily writing tier
13o-series / deliberative reasoners
マップ優先度: 88
Explicit multi-step reasoning route for hard second-pass work
14Claude Haiku / fast tier
マップ優先度: 87
Low-latency triage tier used as cheap front filter
15Phi-4 / small strong on-device
マップ優先度: 86
On-device/edge small models for privacy and offline
16Cohere Command / Embed / Rerank
マップ優先度: 85
Still common specialized components for enterprise RAG stacks
17xAI image/video generation stack
マップ優先度: 84
Same-vendor multimodal gen stack beside Grok chat
18OpenAI Images 2.0 / GPT Image 2
マップ優先度: 83
High-res controllable product-image path used in delivery
19Local GGUF 70B-class (llama.cpp)
マップ優先度: 82
Fully offline private inference baseline for air-gapped envs
20Command R+ / retrieval-first LLMs
マップ優先度: 81
Retrieval-first enterprise QA family still on buy lists
Cover for エージェント枠組み
カバー · Grok Imagine

エージェント枠組み

システムとして出荷するオーケストレーション · top 20 · as_of 2026-07-22

情報源 / 公開API
#名称地図に載る理由
1LangGraph
採用度: 100
Stateful recoverable graphs; a de-facto complex-agent standard
2Claude Agent SDK / Claude Code harness
採用度: 99
Native coding-agent harness with hooks/skills
3OpenAI Agents SDK
採用度: 98
Hosted tool-loop with MCP integration path
4MCP (Model Context Protocol)
採用度: 97
Tool protocol standard across IDEs/runtimes
5CrewAI
採用度: 96
Role-based multi-agent team templates
6AutoGen / AG2
採用度: 95
Multi-agent conversational collaboration patterns
7Semantic Kernel
採用度: 94
Enterprise plugin orchestration (.NET/Python)
8LlamaIndex Workflows
採用度: 93
Data/index-centric agent pipelines
9Dify
採用度: 92
Low-code agent app build and publish
10n8n + AI nodes
採用度: 91
Workflow automation with embedded LLM steps
11Temporal + LLM workers
採用度: 90
Durable long-running agent jobs
12Mastra
採用度: 89
TypeScript-native agent framework
13PydanticAI
採用度: 88
Typed Python agent tooling
14Cursor Agent mode
採用度: 87
In-IDE multi-file agent UX default
15Codex CLI
採用度: 86
Terminal-side OpenAI coding agent
16Aider
採用度: 85
Git-diff-aware CLI pair programmer
17Browser-use / computer-use stacks
採用度: 84
GUI/browser action agents
18Haystack
採用度: 83
Composable NLP/RAG/agent pipelines
19OpenClaw / Hermes-class harnesses
採用度: 82
Personal multi-tool agent OS direction
20Smolagents
採用度: 81
Minimal Hugging Face agent runtime
Cover for コーディングエージェントとIDE
カバー · Grok Imagine

コーディングエージェントとIDE

日々の出荷に使う面 · top 20 · as_of 2026-07-22

情報源 / 公開API
#名称地図に載る理由
1Claude Code
出荷頻度: 100
Deep-repo agent with hooks/skills; strong multi-file reasoning
2Codex CLI / VS Code agent
出荷頻度: 99
Terminal-first autonomous coding; top Terminal-Bench surface
3Cursor
出荷頻度: 98
Mainstream AI IDE default; highest daily multi-file edit polish
4GitHub Copilot Workspace / agent
出荷頻度: 97
PR-native agent with CI; common enterprise procurement default
5Windsurf / Devin Desktop
出荷頻度: 96
Cognition-consolidated agent IDE surface; async ticket-style work
6Grok Build
出荷頻度: 95
xAI coding CLI/build surface; fastest-rising newcomer
7OpenCode
出荷頻度: 94
Open multi-provider coding agent; leading MIT/star open entry
8Gemini CLI
出荷頻度: 93
Google terminal coding agent; free frontier-model entry
9Aider
出荷頻度: 92
Pure CLI + git workflow camp
10Continue.dev
出荷頻度: 91
Open IDE extension hub
11Zed AI
出荷頻度: 90
Performance-first editor AI
12JetBrains AI Assistant
出荷頻度: 89
Embedded in JetBrains ecosystem
13Amazon Q Developer
出荷頻度: 88
AWS-centric code agent
14Cline / Roo Code
出荷頻度: 87
Open VS Code agent fork family
15Replit Agent
出荷頻度: 86
Cloud IDE full-stack agent
16v0 / Lovable class
出荷頻度: 85
UI generation to runnable frontends
17Bolt / StackBlitz agent
出荷頻度: 84
In-browser full-stack generation
18Devin-class autonomous eng agents
出荷頻度: 83
Highly autonomous async eng-agent category
19SWE-agent patterns
出荷頻度: 82
Research-to-prod coding loop patterns
20Sourcegraph Cody
出荷頻度: 81
Code-graph-aware assistant
Cover for RAG・記憶・知識
カバー · Grok Imagine

RAG・記憶・知識

検索とセッション横断メモリ · top 20 · as_of 2026-07-22

情報源 / 公開API
#名称地図に載る理由
1LlamaIndex
成熟度: 100
A default data framework for RAG apps
2LangChain / LangGraph retrievers
成熟度: 99
Broadest connector and chain retrieval ecosystem
3Pinecone
成熟度: 98
Managed vector DB common buy default
4Qdrant
成熟度: 97
Open high-performance vector store
5Weaviate
成熟度: 96
Hybrid search with modules
6pgvector
成熟度: 95
Vectors inside Postgres for existing PG stacks
7Chroma
成熟度: 94
Local-first vector store for prototypes
8Elasticsearch vector
成熟度: 93
Search+vector hybrid upgrade path
9Voyage / Cohere embeddings
成熟度: 92
High-quality embedding APIs
10Docling / Unstructured
成熟度: 91
Document parse-to-chunk entry
11Firecrawl / web extract
成熟度: 90
Web to clean markdown extraction
12GraphRAG patterns
成熟度: 89
Entity-graph retrieval patterns
13Mem0 / agent memory libs
成熟度: 88
Cross-session agent memory libraries
14Azure AI Search
成熟度: 87
Microsoft enterprise hybrid search
15Vertex AI Search
成熟度: 86
GCP enterprise RAG
16NotebookLM class
成熟度: 85
Source-grounded notebooks
17Obsidian + AI plugins
成熟度: 84
Personal knowledge base + AI
18Zep / conversation memory
成熟度: 83
Conversation long-term memory services
19Code graphs (Tree-sitter / codegraph)
成熟度: 82
Structural code retrieval vs pure vectors
20Haystack pipelines
成熟度: 81
Production NLP/RAG pipelines
Cover for 画像生成(現行)
カバー · Grok Imagine

画像生成(現行)

2026の主要パス · top 20 · as_of 2026-07-22

情報源 / 公開API
#名称地図に載る理由
1ChatGPT Images 2.0 / GPT Image 2
出荷可用性: 100
Delivery-oriented high-res controllable images; stronger readable text
2Grok / Imagine
出荷可用性: 99
xAI image gen; usable style-reference edits
3Flux family
出荷可用性: 98
Open-weight photoreal main line
4Midjourney v7 class
出荷可用性: 97
Aesthetic still default
5Ideogram
出荷可用性: 96
Strong text-in-image layout
6Imagen 3/4
出荷可用性: 95
Google product image stack
7Stable Diffusion 3.x
出荷可用性: 94
Open ecosystem baseline
8Adobe Firefly
出荷可用性: 93
Brand-safe commercial licensing narrative
9Recraft
出荷可用性: 92
Design-system / vector-ish outputs
10Leonardo
出荷可用性: 91
Game/asset-oriented generation
11ControlNet stacks
出荷可用性: 90
Pose/depth/edge control
12IP-Adapter / identity ref
出荷可用性: 89
Character identity consistency
13Kolors / CN diffusion
出荷可用性: 88
Chinese aesthetic open routes
14Playground class
出荷可用性: 87
Design-friendly UI generation
15Civitai community models
出荷可用性: 86
Community fine-tune marketplace
16SDXL Lightning / turbo
出荷可用性: 85
Fast preview generation
17Topaz / RealESRGAN upscalers
出荷可用性: 84
Post-gen upscale finish
18Background removal models
出荷可用性: 83
E-commerce cutout necessity
19Inpaint/outpaint toolchains
出荷可用性: 82
Local repaint workflows
20DALL·E lineage integrations
出荷可用性: 81
Legacy integrations still inside many products
Cover for 動画とモーション
カバー · Grok Imagine

動画とモーション

T2V/I2Vとトーキングヘッド · top 20 · as_of 2026-07-22

情報源 / 公開API
#名称地図に載る理由
1Google Veo 3.x
利用パス: 100
Google cinematic T2V/I2V main line; unified multimodal stack
2Sora class
利用パス: 99
High-end narrative T2V category (product entry still shifting)
3Kling
利用パス: 98
Common CN-market video generation
4Runway Gen-3/4
利用パス: 97
Creator video tool stack
5Luma Dream Machine
利用パス: 96
Accessible cinematic T2V
6Pika
利用パス: 95
Social short-form oriented
7Hailuo / MiniMax video
利用パス: 94
CN product video models
8Seedance / ByteDance video
利用パス: 93
ByteDance motion/video pipelines
9Vidu class
利用パス: 92
Multi-subject video
10HeyGen
利用パス: 91
Avatar presenter videos
11Synthesia
利用パス: 90
Enterprise training avatars
12ElevenLabs + lipsync stacks
利用パス: 89
Voice-driven lipsync stacks
13Hedra / talking-head
利用パス: 88
Portrait talking-head video
14CapCut AI suite
利用パス: 87
Mass-market edit + gen
15Descript
利用パス: 86
Transcript-first video edit
16Remotion + AI assets
利用パス: 85
Code-driven video
17HyperFrames / card video
利用パス: 84
Card short-video templates
18Adobe Firefly Video
利用パス: 83
Adobe brand-safe motion
19Viggle / character motion
利用パス: 82
Character motion transfer
20Open-Sora / CogVideo class
利用パス: 81
Open video research routes
Cover for 音声とオーディオ
カバー · Grok Imagine

音声とオーディオ

TTS/STT/音楽/ボイスエージェント · top 20 · as_of 2026-07-22

情報源 / 公開API
#名称地図に載る理由
1ElevenLabs
本番採用: 100
High-quality TTS/clone default
2OpenAI Realtime / TTS
本番採用: 99
Low-latency voice-agent path
3Whisper family
本番採用: 98
Open STT baseline
4Deepgram
本番採用: 97
Production STT APIs
5AssemblyAI
本番採用: 96
STT plus speech understanding
6Cartesia
本番採用: 95
Realtime voice generation
7Azure Speech
本番採用: 94
Enterprise STT/TTS
8Google Cloud Speech
本番採用: 93
GCP speech stack
9Amazon Polly / Transcribe
本番採用: 92
AWS default speech
10Suno
本番採用: 91
Mass-market music generation
11Udio
本番採用: 90
Music generation competitor
12Play.ht / PlayAI
本番採用: 89
Voice product stack
13Resemble AI
本番採用: 88
Enterprise voice clone
14Descript Studio Sound
本番採用: 87
Podcast denoise
15FunASR / SenseVoice
本番採用: 86
Chinese open STT
16Fish Audio / CN TTS
本番採用: 85
Chinese natural TTS
17Coqui / open TTS
本番採用: 84
Open voice models
18WebRTC realtime voice bots
本番採用: 83
Telephony-grade voice bots
19Podcast AI digests
本番採用: 82
Audio-to-summary pipelines
20Audo / cleanup tools
本番採用: 81
Audio cleanup tools
Cover for 評価・可観測・ガードレール
カバー · Grok Imagine

評価・可観測・ガードレール

計測・追跡・制約 · top 20 · as_of 2026-07-22

情報源 / 公開API
#名称地図に載る理由
1LangSmith
運用準備: 100
LangChain-family tracing and evals
2Langfuse
運用準備: 99
Open LLM observability
3Phoenix / Arize
運用準備: 98
ML+LLM monitoring
4Braintrust
運用準備: 97
Eval and logging platform
5Promptfoo
運用準備: 96
Prompt/regression testing
6DeepEval
運用準備: 95
Python eval framework
7RAGAS
運用準備: 94
RAG quality metrics
8Helicone
運用準備: 93
Gateway-layer observability
9Weights & Biases Weave
運用準備: 92
Experiment + LLM traces
10OpenTelemetry GenAI
運用準備: 91
Standardized tracing
11Guardrails AI
運用準備: 90
Output validators
12NeMo Guardrails
運用準備: 89
Dialog policy rails
13Llama Guard (Meta)
運用準備: 88
Safety classifier models
14Presidio / PII redaction
運用準備: 87
Privacy redaction
15LiteLLM
運用準備: 86
Unified model proxy
16Portkey
運用準備: 85
Gateway and fallback routing
17Humanloop
運用準備: 84
Prompt management
18Sentry
運用準備: 83
App+error observability (incl. AI integrations)
19Garak (NVIDIA)
運用準備: 82
LLM adversarial/red-team testing
20OpenAI Evals
運用準備: 81
Custom task eval patterns
Cover for 推論インフラ
カバー · Grok Imagine

推論インフラ

モデルが実際に動く場所 · top 20 · as_of 2026-07-22

情報源 / 公開API
#名称地図に載る理由
1vLLM
配備現実: 100
Default open high-throughput serving
2TensorRT-LLM
配備現実: 99
NVIDIA-optimized inference
3Ollama
配備現実: 98
Local model UX default
4llama.cpp / GGUF
配備現実: 97
CPU/GPU local quantization
5MLX
配備現実: 96
Apple Silicon inference
6Fireworks AI
配備現実: 95
Fast hosted open models
7Together AI
配備現実: 94
Open-model cloud
8Groq
配備現実: 93
Ultra-low-latency LPU inference
9AWS Bedrock
配備現実: 92
Enterprise multi-model hosted API
10Azure OpenAI / AI Foundry
配備現実: 91
Microsoft enterprise model platform
11Vertex AI
配備現実: 90
GCP model platform
12Hugging Face TGI / Inference
配備現実: 89
HF serving path
13Modal
配備現実: 88
Serverless GPU jobs
14Replicate
配備現実: 87
Model-as-API marketplace
15Cloudflare Workers AI
配備現実: 86
Edge inference
16NVIDIA NIM
配備現実: 85
Packaged enterprise inference microservices
17SGLang
配備現実: 84
Fast open serving
18Kubernetes
配備現実: 83
DIY production orchestration baseline
19SkyPilot
配備現実: 82
Multi-cloud GPU cluster orchestration
20Anyscale / Ray
配備現実: 81
Distributed train and serving
Cover for AIリーダー地図(2026)
カバー · Grok Imagine

AIリーダー地図(2026)

フォロワー数ではなく議題を定義する人 · top 20 · as_of 2026-07-22

情報源 / 公開API
#名称地図に載る理由
1Demis Hassabis (Google DeepMind)
議題影響力: 100
Science+product dual narrative; Alpha lineage + Gemini control plane
2Dario Amodei (Anthropic)
議題影響力: 99
Safety/interpretability commercial flagship; Claude line
3Sam Altman (OpenAI)
議題影響力: 98
Platform distribution + capital narrative center; ChatGPT default
4Yann LeCun (Meta FAIR / AMI Labs)
議題影響力: 97
Open-source + world-model public debate bridge academia↔industry
5Jensen Huang (NVIDIA)
議題影響力: 96
De-facto tax rate of the compute supply chain
6Ilya Sutskever (SSI)
議題影響力: 95
Scaling+alignment legacy and next-gen SSI narrative
7Mustafa Suleyman (Microsoft AI)
議題影響力: 94
Consumer+enterprise dual stack; Copilot execution authority
8Fei-Fei Li
議題影響力: 93
Human-centered AI + vision foundations public voice
9Geoffrey Hinton
議題影響力: 92
Godfather-layer public risk voice; high policy weight
10Yoshua Bengio
議題影響力: 91
Safety/governance academic agenda-setting
11Andrej Karpathy
議題影響力: 90
Engineering education + product intuition mass amplifier
12Satya Nadella (Microsoft)
議題影響力: 89
CEO who ships frontier models into enterprise distribution
13Mark Zuckerberg (Meta)
議題影響力: 88
Open-weight + social distribution free-tier strategy
14Elon Musk (xAI)
議題影響力: 87
Realtime X data + Grok product alternate narrative center
15Noam Shazeer
議題影響力: 86
Transformer co-author → consumer-agent product path
16Jared Kaplan (Anthropic)
議題影響力: 85
Scaling-laws→alignment productization research leadership
17Jim Fan (NVIDIA)
議題影響力: 84
Public flagship for embodied/game agents research
18Percy Liang (Stanford CRFM)
議題影響力: 83
HELM-class measurable-trust eval infrastructure
19Andrew Ng
議題影響力: 82
Education + applied-AI global distribution layer
20Oriol Vinyals (DeepMind)
議題影響力: 81
Sequence models + multi-agent research leadership
Cover for トップラボ / フロンティア組織
カバー · Grok Imagine

トップラボ / フロンティア組織

モデル・論文・スタック — 公式X付き · top 20 · as_of 2026-07-22

情報源 / 公開API
#名称地図に載る理由
1Google DeepMind
システム影響: 100
Science breakthrough + Gemini; Search/Cloud distribution
2OpenAI
システム影響: 99
ChatGPT default entry; API + agent platform center
3Anthropic
システム影響: 98
Claude long-horizon/safety stack; enterprise coding narrative
4Meta FAIR / Superintelligence Labs
システム影響: 97
Llama open-weight default base; research→community flywheel
5Microsoft Research / Microsoft AI
システム影響: 96
Copilot distribution + Azure hosting; enterprise buy path
6xAI
システム影響: 95
Grok + X realtime surface; Imagine multimodal
7NVIDIA Research / AI
システム影響: 94
Compute + NIM inference; supply choke point
8Amazon Science / AWS AI
システム影響: 93
Bedrock multi-model shelf; default-visible in AWS accounts
9Mistral AI
システム影響: 92
EU-compliance open/API dual track
10DeepSeek
システム影響: 91
Open reasoning cost shock; private-deploy reference
11Alibaba Qwen / Tongyi
システム影響: 90
Bilingual + tool-use ecosystem; open and cloud
12Moonshot / Kimi
システム影響: 89
Strong Chinese long-context product
13Cohere
システム影響: 88
Enterprise RAG/retrieval + Command line
14Hugging Face
システム影響: 87
Model/dataset/Spaces distribution hub
15Allen Institute for AI (AI2)
システム影響: 86
Open science path incl. OLMo-class models
16Stanford CRFM
システム影響: 85
Eval×policy academic hub (HELM)
17Stability AI
システム影響: 84
Image-gen open-weight and API product node
18Black Forest Labs (FLUX)
システム影響: 83
FLUX image-gen research/product line
19Perplexity
システム影響: 82
Retrieval-native chat product; active public eng narrative
20Runway Research
システム影響: 81
Video-gen research+product; Gen-series public reference
Cover for 里程標論文 / 思想アンカー
カバー · Grok Imagine

里程標論文 / 思想アンカー

各行が正規論文リンク · top 20 · as_of 2026-07-22

情報源 / 公開API
#名称地図に載る理由
1Attention Is All You Need (Vaswani et al.)
思想遺産: 100
Origin of the default Transformer architecture
2Scaling Laws / Kaplan et al.
思想遺産: 99
Start of predictable compute-data-params narrative
3GPT-3 / few-shot learners
思想遺産: 98
Public proof that large pretrain yields general ability
4InstructGPT / RLHF
思想遺産: 97
Productization path aligning to human preference
5Constitutional AI (Anthropic)
思想遺産: 96
Discussable blueprint for principle-based alignment
6Chain-of-Thought prompting
思想遺産: 95
Made reasoning an explicit controllable interface
7ReAct (reason + act)
思想遺産: 94
Default mental model for tool-using agents
8LLaMA (Meta)
思想遺産: 93
Turn where open weights became a default base
9Chinchilla compute-optimal
思想遺産: 92
Efficiency correction: data vs parameters
10Diffusion / DDPM family
思想遺産: 91
Image-gen main path shift GAN→diffusion
11CLIP (OpenAI)
思想遺産: 90
Vision-language shared representation into products
12AlphaFold (DeepMind)
思想遺産: 89
Undeniable science-AI achievement sample
13FlashAttention
思想遺産: 88
Feasibility jump for long-context train/serve
14Mixture-of-Experts / Switch Transformers
思想遺産: 87
Sparse activation scaling became mainstream
15DPO (Direct Preference Optimization)
思想遺産: 86
Widely adopted simplified alignment training path
16Toolformer
思想遺産: 85
Early clear demo of models learning to call APIs
17Segment Anything (Meta)
思想遺産: 84
Vision foundation model train-once multi-downstream template
18Whisper (OpenAI)
思想遺産: 83
One default open baseline for speech recognition
19BERT (Devlin et al.)
思想遺産: 82
Prior-gen turn: bidirectional pretrain as NLP industrial default
20LoRA / PEFT family
思想遺産: 81
Cheap adapters made private fine-tune a default move
Cover for 買い手が実際に引用するベンチ
カバー · Grok Imagine

買い手が実際に引用するベンチ

各行が公式ボード/リポジトリ · top 20 · as_of 2026-07-22

情報源 / 公開API
#名称地図に載る理由
1LM Arena / Chatbot Arena Elo
意思決定引用率: 100
Public arena perceived quality; often cited in sales (arena.ai)
2SWE-bench / SWE-bench Verified
意思決定引用率: 99
Real software tasks; coding-agent purchase yardstick
3HumanEval
意思決定引用率: 98
Classic code-gen yardstick (OpenAI)
4LiveCodeBench
意思決定引用率: 97
Continuously updated code-gen board
5MMLU
意思決定引用率: 96
Multi-subject knowledge default (watch saturation)
6GPQA
意思決定引用率: 95
Hard science QA; real-reasoning cross-check
7MATH dataset
意思決定引用率: 94
Math-reasoning strength cross-check
8HELM (Stanford CRFM)
意思決定引用率: 93
Multi-scenario reproducible academic eval framework
9Artificial Analysis
意思決定引用率: 92
Quality×speed×price three-axis purchase view
10Open LLM Leaderboard (HF)
意思決定引用率: 91
Open-weight public comparison entry
11τ-bench (tool-agent)
意思決定引用率: 90
Tool-use reliability; critical for agent products
12WebArena
意思決定引用率: 89
Browser-agent task completion
13MMMU benchmark
意思決定引用率: 88
Multimodal understanding purchase view
14VBench
意思決定引用率: 87
Video-gen perception and consistency
15Common Voice (ASR data)
意思決定引用率: 86
Public ASR corpus/entry point
16RAGAS
意思決定引用率: 85
RAG faithfulness and retrieval quality
17AgentBoard
意思決定引用率: 84
Multi-agent task composite board
18BIG-bench
意思決定引用率: 83
Prior-gen composite ability (legacy yardstick)
19MLPerf Inference
意思決定引用率: 82
Hardware/serving throughput industry yardstick
20LMSYS Chatbot Arena paper
意思決定引用率: 81
Arena methodology paper (how to read Elo)
Cover for Product Hunt · AIプロダクト索引
カバー · Grok Imagine

Product Hunt · AIプロダクト索引

ローンチ面でよく名指しされるAI製品——各行ホームリンク · top 20 · as_of 2026-07-22

情報源 / 公開API
#名称地図に載る理由
1ChatGPT
流通ヒート: 100
Default chat entry; launch-narrative center
2Claude
流通ヒート: 99
Long-form/coding chat product surface
3Cursor
流通ヒート: 98
AI IDE path from launch to mainstream
4Perplexity
流通ヒート: 97
Retrieval-native answer engine; strong distribution narrative
5v0 (Vercel)
流通ヒート: 96
UI-gen→deployable launch product template
6Lovable
流通ヒート: 95
NL→full-stack app PH-hot product shape
7Bolt.new
流通ヒート: 94
In-browser one-shot full-stack prototypes
8Gamma
流通ヒート: 93
AI decks/docs launch tool
9Notion AI
流通ヒート: 92
Workspace-native writing/Q&A
10Raycast AI
流通ヒート: 91
Launcher-native AI command layer
11Linear
流通ヒート: 90
Eng project management + AI assist
12ElevenLabs
流通ヒート: 89
Voice gen/clone product surface
13Suno
流通ヒート: 88
Mass-market music-gen launch breakout shape
14Runway
流通ヒート: 87
Creator video-gen tool stack
15Ideogram
流通ヒート: 86
Text-in-image friendly image product
16Replicate
流通ヒート: 85
Model-as-API marketplace; dev distribution
17Hugging Face
流通ヒート: 84
Model/Spaces community distribution hub
18Descript
流通ヒート: 83
Transcript-first A/V product
19Midjourney
流通ヒート: 82
Aesthetic still-gen mass entry (main site often bot-walled)
20Product Hunt AI topic
流通ヒート: 81
Master index of AI launches (topic page)
Cover for Apple App Store · AIアプリ索引
カバー · Grok Imagine

Apple App Store · AIアプリ索引

iOSの主要AI入口——各行App Storeリンク · top 20 · as_of 2026-07-22

情報源 / 公開API
#名称地図に載る理由
1ChatGPT
ストア可視性: 100
Default consumer chat entry
2Claude by Anthropic
ストア可視性: 99
Long-form/coding assistant on iOS
3Google Gemini
ストア可視性: 98
Google multimodal assistant entry
4Microsoft Copilot
ストア可視性: 97
Microsoft assistant + OS integration narrative
5Grok
ストア可視性: 96
xAI assistant; X-ecosystem distribution
6Perplexity
ストア可視性: 95
Retrieval-native Q&A mobile entry
7Character.AI
ストア可視性: 94
Character-chat consumer breakout shape
8PhotoRoom AI Photo Editor
ストア可視性: 93
E-commerce/cutout AI photo edit
9Remini – AI Photo Enhancer
ストア可視性: 92
Photo upscale/enhance breakout
10Lensa: AI Photo & Video Editor
ストア可視性: 91
Portrait stylization AI editor
11FaceApp: Face Editor
ストア可視性: 90
Face-transform AI filters
12CapCut – Video Editor
ストア可視性: 89
Short-video edit + AI features
13Canva: AI Photo & Video Editor
ストア可視性: 88
Design templates + generative features
14Grammarly Keyboard
ストア可視性: 87
Writing-assistant keyboard layer
15Duolingo
ストア可視性: 86
Language learning + AI conversation
16Speak: Language Learning
ストア可視性: 85
Speaking-practice AI coach
17Notion
ストア可視性: 84
Notes/wiki + Notion AI
18starryai – AI Art Generator
ストア可視性: 83
Text-to-image consumer entry
19Wondershare Wonder AI Art
ストア可視性: 82
AI art generation consumer tool
20Microsoft 365 (Copilot surface)
ストア可視性: 81
Office suite Copilot distribution surface
Cover for Google Play · AIアプリ索引
カバー · Grok Imagine

Google Play · AIアプリ索引

Androidの主要AI入口——各行Play詳細 · top 20 · as_of 2026-07-22

情報源 / 公開API
#名称地図に載る理由
1ChatGPT
ストア可視性: 100
Default consumer chat entry
2Claude by Anthropic
ストア可視性: 99
Long-form/coding assistant on Android
3Google Gemini
ストア可視性: 98
Google multimodal assistant entry
4Microsoft Copilot
ストア可視性: 97
Microsoft assistant Android entry
5Grok
ストア可視性: 96
xAI assistant; X-ecosystem distribution
6Perplexity
ストア可視性: 95
Retrieval-native Q&A mobile entry
7Character.AI
ストア可視性: 94
Character-chat consumer breakout shape
8PhotoRoom
ストア可視性: 93
E-commerce/cutout AI photo edit
9Remini
ストア可視性: 92
Photo upscale/enhance breakout
10Lensa
ストア可視性: 91
Portrait stylization AI editor
11FaceApp
ストア可視性: 90
Face-transform AI filters
12CapCut
ストア可視性: 89
Short-video edit + AI features
13Canva
ストア可視性: 88
Design templates + generative features
14Grammarly Keyboard
ストア可視性: 87
Writing-assistant keyboard layer
15Duolingo
ストア可視性: 86
Language learning + AI conversation
16Notion
ストア可視性: 85
Notes/wiki + Notion AI
17Microsoft 365 (Office)
ストア可視性: 84
Office suite Copilot distribution surface
18Background Eraser / remove.bg class
ストア可視性: 83
Background-removal AI tools
19starryai (Play search)
ストア可視性: 82
Text-to-image consumer entry (search index)
20Speak-class language AI (Play search)
ストア可視性: 81
Speaking AI coach category index
Cover for Xホットトピック · AI
カバー · Grok Imagine

Xホットトピック · AI

X上の主要AI議題——各行検索/ハッシュタグ · top 20 · as_of 2026-07-22

情報源 / 公開API
#名称地図に載る理由
1AI Agents
議題ヒート: 100
Autonomous agents / tool-use as main narrative
2Claude / Anthropic
議題ヒート: 99
Long-form, coding, and safety narrative center
3ChatGPT / OpenAI
議題ヒート: 98
Default chat entry and release-cadence talk
4Grok / xAI
議題ヒート: 97
X-native assistant and model-release talk
5Gemini / Google AI
議題ヒート: 96
Multimodal and search-integration agenda
6DeepSeek
議題ヒート: 95
Open-weight / efficient-training contrast narrative
7Open-weight LLMs
議題ヒート: 94
Qwen / Kimi / Mistral open-weight wave
8MCP (Model Context Protocol)
議題ヒート: 93
Tool-interop standard as infrastructure talk
9Cursor / AI IDE
議題ヒート: 92
In-editor coding-agent product surface
10AI Video Generation
議題ヒート: 91
Sora / Kling / Seedance-class video gen
11Image Gen / Flux
議題ヒート: 90
T2I aesthetics and local-model discussion
12Voice AI / Realtime
議題ヒート: 89
Realtime voice assistants and call automation
13AI Memory / Personalization
議題ヒート: 88
Memory and personalization as product competition
14Local / On-device AI
議題ヒート: 87
On-device and local inference feasibility
15AI Jobs / Automation
議題ヒート: 86
Jobs and knowledge-labor automation agenda
16AI Safety (near-term)
議題ヒート: 85
Near-term safety / bio / cyber risk talk
17Evals / Arena
議題ヒート: 84
Public arena and purchase-eval yardsticks
18RAG / Knowledge
議題ヒート: 83
RAG and enterprise knowledge bases
19AI Regulation
議題ヒート: 82
Policy, copyright, and deployment compliance
20Multimodal / World Models
議題ヒート: 81
Multimodal and world-model frontier narrative
Cover for Reddit · AIホットトピック
カバー · Grok Imagine

Reddit · AIホットトピック

AIコミュニティ主要議題——subreddit/hot検索 · top 20 · as_of 2026-07-22

情報源 / 公開API
#名称地図に載る理由
1r/MachineLearning · hot
コミュニティヒート: 100
Papers, methods, and review-culture main community
2r/LocalLLaMA · hot
コミュニティヒート: 99
Local models / quant / hardware practice
3r/OpenAI · hot
コミュニティヒート: 98
ChatGPT/API product and policy talk
4r/ClaudeAI · hot
コミュニティヒート: 97
Claude product use and comparisons
5r/ChatGPT · hot
コミュニティヒート: 96
Mass-market chat product experience feedback
6r/singularity · hot
コミュニティヒート: 95
Long-run capability curve and social-impact narrative
7r/artificial · hot
コミュニティヒート: 94
General AI news and discussion
8r/StableDiffusion · hot
コミュニティヒート: 93
Open image-gen workflows
9r/midjourney · hot
コミュニティヒート: 92
Aesthetic stills and prompt culture
10r/LangChain · hot
コミュニティヒート: 91
Orchestration / tool-chain engineering practice
11r/Oobabooga · hot
コミュニティヒート: 90
Local WebUI inference stack
12r/LocalAI · hot
コミュニティヒート: 89
Self-hosted inference and API-compat layers
13r/PromptEngineering · hot
コミュニティヒート: 88
Prompt and system-instruction craft
14r/MLOps · hot
コミュニティヒート: 87
Train/deploy/monitor productionization
15r/computervision · hot
コミュニティヒート: 86
Vision models and applications
16Coding agents (Reddit search)
コミュニティヒート: 85
Cursor/Copilot/agent coding heat
17AI video gen (Reddit search)
コミュニティヒート: 84
Video-gen models and tools discussion
18RAG / memory (Reddit search)
コミュニティヒート: 83
RAG and long-term memory practice
19Open-weight models (Reddit search)
コミュニティヒート: 82
Open-weight releases and benchmark contrast
20AI jobs / automation (Reddit search)
コミュニティヒート: 81
Jobs and workflow-replacement talk
Cover for GitHub · AIリポジトリ索引
カバー · Grok Imagine

GitHub · AIリポジトリ索引

主要OSS AIリポジトリ——各行GitHub · top 20 · as_of 2026-07-22

情報源 / 公開API
#名称地図に載る理由
1ollama/ollama
エコ可視性: 100
Local model runtime; Docker-like one-shot pull
2ggml-org/llama.cpp
エコ可視性: 99
CPU/GPU local quantized inference base
3huggingface/transformers
エコ可視性: 98
Default model load/train library
4vllm-project/vllm
エコ可視性: 97
High-throughput open serving
5sgl-project/sglang
エコ可視性: 96
Fast structured-generation serving
6langchain-ai/langchain
エコ可視性: 95
Default app orchestration ecosystem
7langchain-ai/langgraph
エコ可視性: 94
Stateful agent-graph runtime
8openai/codex
エコ可視性: 93
Terminal/repo coding-agent reference
9Aider-AI/aider
エコ可視性: 92
Git-native pair-programming agent
10All-Hands-AI/OpenHands
エコ可視性: 91
Open software-engineering agent platform
11cline/cline
エコ可視性: 90
IDE-side autonomous coding agent
12continuedev/continue
エコ可視性: 89
Open IDE autocomplete/agent extension
13modelcontextprotocol/servers
エコ可視性: 88
MCP official/community server collection
14comfyanonymous/ComfyUI
エコ可視性: 87
Node-based image/video gen workflows
15open-webui/open-webui
エコ可視性: 86
Self-hosted ChatGPT-style frontend
16langgenius/dify
エコ可視性: 85
Visual LLM app/workflow platform
17n8n-io/n8n
エコ可視性: 84
Workflow automation + AI nodes
18deepseek-ai/DeepSeek-V3
エコ可視性: 83
Open-weight frontier model release anchor
19QwenLM/Qwen3
エコ可視性: 82
Qwen open-model series entry
20microsoft/autogen
エコ可視性: 81
Multi-agent conversation framework
Cover for AI Skills · エコ索引
カバー · Grok Imagine

AI Skills · エコ索引

Agent/Claude Skillsと配信面——docs/repo · top 20 · as_of 2026-07-22

情報源 / 公開API
#名称地図に載る理由
1Anthropic Agent Skills (docs)
配信可視性: 100
Official skill format and usage contract
2anthropics/skills
配信可視性: 99
Official public skill sample repo
3vercel-labs/agent-skills
配信可視性: 98
Engineering-oriented agent skills pack (Vercel)
4awesome-claude-skills
配信可視性: 97
Curated Claude skills list
5AgentSkill.sh
配信可視性: 96
Third-party skill discovery directory
6ClawHub
配信可視性: 95
OpenClaw-family skill distribution surface
7modelcontextprotocol/servers
配信可視性: 94
MCP servers as pluggable tool-skill layer
8GitHub Skills
配信可視性: 93
Official GitHub interactive learning skills
9openai/codex
配信可視性: 92
Codex runtime + extension/skill surface
10Aider-AI/aider
配信可視性: 91
In-repo coding-skill habit reference
11cline/cline
配信可視性: 90
IDE agent skill/tool composition surface
12continuedev/continue
配信可視性: 89
Open IDE extension skill ecosystem
13crewAIInc/crewAI
配信可視性: 88
Multi-role agent skill orchestration
14microsoft/autogen
配信可視性: 87
Conversational multi-agent skill composition
15langchain-ai/langgraph
配信可視性: 86
Graph state machine for reusable sub-skills
16All-Hands-AI/OpenHands
配信可視性: 85
Software-engineering agent skill platform
17firecrawl/firecrawl
配信可視性: 84
Web-crawl skill (agent data ingress)
18BerriAI/litellm
配信可視性: 83
Multi-model routing skill middleware
19langgenius/dify
配信可視性: 82
Visual workflow as composable skill canvas
20n8n-io/n8n
配信可視性: 81
Automation node skill library
Live secondary evidence (GitHub · not editorial rank)

snapshot 2026-07-22T11:13:50.347893+00:00 · secondary only