SceneXplain
3.4GPT-4-powered computer vision writes rich, accurate captions for complex scenes with multiple objects and interactions – define your own JSON schema for structured output.
-
-
Elize.AI
3.3Fully automatic, real-time 3D character animation – face, speech, and body movement generated on the fly based on emotional state, no motion capture needed.
-
Tensorbox
3.0Fixes logic and advanced errors in Python and JavaScript, then helps you train and deploy your own custom GPT model – a no-code interface for beginners, a single API call for developers.
-
SnapEdit
3.4One-touch AI object removal, background swaps from 100+ templates, batch process 50 photos at once – 20M+ users, 300M+ photos edited monthly.
-
finbots.ai
3.4Build a credit scorecard in a day instead of 9-12 months – creditX makes lending decisions in under 0.03 seconds, reporting 20%+ higher approvals and 15%+ lower losses.
-
Moveo.AI
3.4AI agents work across voice, chat, email, and WhatsApp with memory that persists across every interaction – reported 75% lower call center costs and 90% automation rates for banks and insurers.
-
AILab Tools
3.2Free web-based portrait enhancement, background removal, and virtual try-on, plus 50+ commercial APIs for the same – files auto-deleted within 24 hours.
-
Bitskout
3.3Show it a few examples and Bitskout AI learns to extract exactly the fields you need from any invoice, CV, or bill of lading – then pushes the data straight into monday.com or Zapier.
-
AppTek.ai
3.4Transcribe, translate, understand, and synthesize speech across dozens of languages – ASR, neural machine translation, LLMs, TTS, and NLP in one enterprise language-AI platform.
-
Backstory
3.4Formerly People.ai – AI reads every email, call, and meeting to score deals against the MEDDPICC framework and flag pipeline risk, answered directly inside Claude, ChatGPT, or Slack.
-
Kadoa
3.4Describe the dataset you need in plain language – AI subagents discover, navigate, extract, and validate it from websites, PDFs, and spreadsheets, with every value traced back to its source.
-
Signapse
3.6Real-time AI sign-language translation for broadcast, transit, and government – generative AI blends a library of qualified sign-language video into natural, flowing translations.
-
Hoop
3.8An AI retention agent that asks subscribers why they're canceling and makes a personalized save offer before the cancellation confirms – built by the team behind Trello.
-
ClipMyApp
3.7Turn one app recording into a week of social content with MCP agent control.
-
Certainly
3.6Agentic AI agents that reason, act, and resolve customer inquiries end-to-end across web, email, SMS, and voice – choose from Claude, Gemini, GPT, or 200+ other models.
-
Diffbot
3.5Computer vision renders and classifies any webpage into clean structured data in under 300ms, then feeds it into a Knowledge Graph spanning billions of people, organizations, and products.
-
Landing AI
3.7Agentic APIs turn unstructured documents into structured, auditable data – proprietary vision-first models process thousands of pages a minute for Barclays, Morgan Stanley, and Intel.
-
Promptitude.io
3.2One dashboard to organize, test, and compare prompts across any AI provider – upload your own docs to ground responses, then share a prompt as an embeddable widget or connect it via API.
-
Defined.ai
3.6The world's largest ethically-sourced AI data marketplace – trusted by Adobe, Amazon, Google, Meta, and OpenAI, custom data collection across 150+ countries and 500+ languages.
-
Removal.AI
3.3Automatically detects the subject in a photo and removes the background in under 3 seconds – batch process up to 1,000 images at once, with API access.
-
Formwise AI
3.3Turn your expert methodology into a branded AI product – guided SmartForms, conversational CoPilots, and voice Experts, on any model (Claude, OpenAI, Gemini, Mistral), no code required.
-
SpeechGPT
2.7A no-code speech recognition and NLU API for building voice-enabled apps – customizable language, accent, and dialect handling for customer service bots and voice search.
-
Syntho
3.3Three ways to generate safe test data – PII masking, rule-based patterns, or full AI-generated synthetic data mimicking real statistical behavior – deployed inside your own environment.
-
Illustroke
3.7Text prompt to stunning vector illustration – 1M+ illustrations created, with an API for building the same generation into your own product.
-
Linkdelta
3.560+ AI writing tools plus code generation (PHP, SQL, Python) across 36 languages – flat EUR tiers from €9/month, no confusing per-word pricing.
-
Sleep.ai
3.3Sleep Sense detects sleep-wake boundaries from your everyday smartphone signals – no wearable needed – built on the world's most comprehensive sleep dataset.
-
DeepGenX
2.9Write a prompt in a Python file, get working code back – CodeGenX runs on GPT-J, a 6-billion-parameter model fine-tuned specifically on open-source Python.
-
Eventual AI
3.4Index every camera on your robot fleet, search for rare failure scenarios in plain language, and auto-curate training datasets – built on the open-source Daft data engine.
-
Neurotechnology
3.836 years of biometric AI R&D – fingerprint, face, iris, voice and palm identification SDKs, deployed in 140+ countries, multiple NIST award wins.
-
Clipping Magic
3.9Auto-Clip AI trained on millions of real-world images (not lab data) removes backgrounds automatically, then a Smart Editor with a hair tool and scalpel handles the hard edges by hand.
-
Capitol AI
3.9No-code AI workflows with governed data access, lineage tracking and full audit trails – built for regulated finance, government and consulting work, not general productivity.
-
DocsBot
3.9Trains on 38+ content sources and deploys across 7,000+ apps via Zapier – now with voice-agent support and deep research combining your knowledge base with live web search.
-
Metaflow
3.9Netflix's open-source Python framework for ML and data-science workflows – free, real production adoption at 23andMe, CNN and Realtor.com.
-
Anam
4.0Photorealistic AI avatars at 180ms average latency, 70+ languages – HIPAA compliant and SOC 2 certified, claims a 44% engagement lift by giving an agent a face.
-
Phonic
4.0Speech-to-speech voice agents at 300ms latency – backed by Lux Capital and founders from Hugging Face, Replit and Modal – built for high-stakes customer calls.
-
Connexun
3.4B.I.R.B.AL., a proprietary AI engine trained on over a million articles, aggregates and classifies real-time news from 20,000+ sources with deep-learning summarization across languages.
-
Vectara
3.7Build and deploy AI agents with always-on hallucination enforcement and context-engineering grounded in your own documents – on-prem, VPC, or SaaS, built for regulated industries.
-
Airy
3.9Real-time AI agents that watch your data streams and act on complex patterns automatically – built on Apache Flink, acquired by Confluent.
-
Logmeal
3.4Point a camera at a plate and get instant nutritional analysis – AI food recognition trained on the world's largest food image dataset, identifying 1,300+ dishes.
-
HandtextAI
3.5Typed text to realistic handwritten pages – your own handwriting as a custom font, or a batch-personalization API for real postcards and letters.
-
Monetizebot
3.4Contextual ad placement inside AI chatbots – a three-sided marketplace for assistant developers, publishers and advertisers, AdSense-style but for conversational AI.
-
ZenAion
3.5Describe a task in conversation and it becomes a workflow that keeps running.
-
iGPT
3.6Rebuilds a full email thread from scattered messages and cites its sources back to them.
-
Vsub
3.2Faceless short-form video in animation styles, with Reddit story automation.
-
ASMR Videos
2.7Veo 3 pointed at one genre, with templates and a permanent sale.
-
PublicAI
3.7Data annotation through a 3.5M-worker network, with validators staking on quality.
-
Qubrid AI
3.7H200s, B200s and serverless inference on an open-source platform.
-
Tabscanner
3.7Receipt OCR as an API, running since 2016 with published per-credit rates.