Inflectiv builds living datasets that AI agents query through API, SDK, or MCP instead of static files. Creators maintain curated records, agents pay per question in credits settled in USDC, and write-back annotations can improve datasets over time.
Interfaze is a model architecture built for deterministic tasks such as OCR, audio understanding, structured data extraction, web scraping, and classification. The vendor publishes token-based pricing with infrastructure included and positions the model for high-accuracy, repeatable outputs compared with general LLMs.
Inworld provides realtime AI infrastructure for consumer apps, including first-party TTS, STT, LLM routing, and inference APIs. Teams integrate voice agents, companions, and learning experiences while choosing on-demand or monthly credit plans.
Isoquant provides optimized inference APIs for open models with published per-token pricing, low latency benchmarks, and OpenAI-compatible chat completion endpoints. Enterprise offerings include workload-specific model tuning, evaluations, and deployment of trained specialists.
Jan is an open-source, privacy-focused ChatGPT alternative that runs 100% offline on your computer. It features an intuitive desktop interface, a local inference engine (Cortex), extensions support, and an OpenAI-compatible local API server for offline AI development.
Jev is TypeSafe’s structured decision model for text and JSON “system one” responses, exposed via playground, batch jobs, and a /v1/systemone API. Accounts receive free input tokens on sign-in, with one-time token packs when usage grows.
Jina AI provides search foundation APIs including Reader for URL-to-markdown, web search, embeddings, rerankers, and MCP access for LLM agents. Developers use r.jina.ai, s.jina.ai, and api.jina.ai endpoints with API keys, rate limits, and optional token top-ups.
Kailash Labs is a video research lab building small, grounded video understanding models and benchmarks. The team converts world video into structured data machines can reason over and offers early SDK access to design partners in robotics, security, industrial, sports, and media.
Kaldi is an open source speech recognition toolkit maintained by the research community. It provides training and decoding recipes for automatic speech recognition experiments, with code hosted on GitHub and documentation on kaldi-asr.org.
Keras is a high-level deep learning API focused on readable, fast-to-debug model code for humans. It ships with Keras 3 multi-backend support, extensive guides, and companion libraries like KerasHub pretrained models and KerasTuner for hyperparameter search.
Kie.ai is a pay-as-you-go API platform that gives developers one API key and one credit balance for third-party image, video, music and language models such as Veo, Kling, Suno, Nano Banana and Gemini, with a web playground for testing.
Kili Technology provides a collaborative data labeling platform for image, video, text, PDF, geospatial, and audio datasets with quality workflows, AI-assisted pre-annotation, and enterprise security certifications. Teams use it to build training, fine-tuning, and evaluation data for ML products.
Kimi API Platform exposes Moonshot AI models—including Kimi K3, K2.7 Code, and K2.6—with long context, tool calling, vision, and official tools such as web search and code execution. Developers authenticate with API keys via OpenAI- or Anthropic-compatible endpoints and pay as they consume tokens.
Kimi K2 is Moonshot AI's open-weight mixture-of-experts model family built for agentic coding, long-context reasoning, and tool use, with releases such as K2.7 Code highlighted on Moonshot's official site. Developers access models through Kimi apps, Moonshot's open platform, and API docs rather than a separate K2 storefront.
Kimi K2.6 is a general-purpose multimodal model on Kimi API Platform supporting text, image, and video input, thinking and non-thinking modes, tool calling, and a 256K-token context window. It targets chat, agent, visual understanding, and complex reasoning workloads billed per token.
Kimpton provides evaluation infrastructure for models and agents, including private environments, behavioral benchmarks, and Koliseum-style live evaluations. Teams run benchmarks via CLI or API, compare checkpoints on long-horizon and multi-agent tasks, and export structured reports.
Klavis AI builds live MCP environments for training and evaluating AI agents on long-horizon tool use. The company targets frontier labs that need realistic worlds, documentation, and open-source tooling rather than a general chat product.
Kong Gateway is Kong Inc.'s API and AI connectivity runtime for securing, routing, and observing traffic across hybrid and multi-cloud environments. Konnect adds a control plane for API management, AI gateways, service mesh, and developer portal capabilities.
Kornia is a non-profit computer vision stack for robotics and spatial AI spanning differentiable PyTorch operators, Rust kernels, SLAM, TensorRT inference, and the Bubbaloop agent runtime for edge devices. The project helps teams prototype in notebooks and deploy the same geometry to Jetson, Raspberry Pi, or Linux robots.
Kubeflow is the open-source Kubernetes-native platform for data, ML, and GenAI workloads, composed of modular projects for training, notebooks, hyperparameter tuning, and model registry. It is community maintained under the CNCF and free to deploy on your own clusters.
Labelbox builds AI training and agentic work environments for frontier labs and enterprises, spanning data labeling, model evaluation, and Recursion-style multi-agent sessions inside business apps. Horizon and Recursion products coordinate specialists that read documents, call tools, and produce graded deliverables.
Not listed
API available
Checked Oct 4, 2026
Showing 360 of 660 tools
Plan facts carry individual sources and check dates. Unknown values are excluded from required filters.