{
  "slug": "call-llm-apis",
  "label": "use LLM inference APIs",
  "blurb": "Hosted and local inference APIs, most of them OpenAI-compatible, that serve open, proprietary and on-device models.",
  "keywords": [
    "llm inference",
    "llm api",
    "inference api",
    "open models",
    "local llm"
  ],
  "categories": [
    "language",
    "infrastructure"
  ],
  "access": [
    "api"
  ],
  "count": 54,
  "official": 35,
  "url": "https://agentswitchboard.dev/tasks/call-llm-apis",
  "browseUrl": "https://agentswitchboard.dev/browse?task=call-llm-apis",
  "agentsUrl": "https://agentswitchboard.dev/api/agents?task=call-llm-apis",
  "agents": [
    {
      "name": "NVIDIA NIM API",
      "slug": "nvidia-nim-api",
      "url": "https://agentswitchboard.dev/agents/nvidia-nim-api",
      "iconUrl": "https://agentswitchboard.dev/icons/nvidia-nim-api.png?v=e1ac122c",
      "description": "Serves open models as NVIDIA-hosted NIM endpoints through an OpenAI-compatible API, with chat, embedding, reranking and image models, for developers prototyping before self-hosting.",
      "provider": "NVIDIA",
      "categories": [
        "language",
        "infrastructure"
      ],
      "accessMethods": [
        "api"
      ],
      "verified": true,
      "official": true,
      "featured": false,
      "sponsored": false,
      "authType": "apiKey",
      "why": "OpenAI-Compatible Chat: Runs chat completions on hosted open models at integrate.api.nvidia.com/v1, with SSE streaming and tool calling."
    },
    {
      "name": "Crusoe Managed Inference",
      "slug": "crusoe-managed-inference",
      "url": "https://agentswitchboard.dev/agents/crusoe-managed-inference",
      "iconUrl": "https://agentswitchboard.dev/icons/crusoe-managed-inference.png?v=08cbe186",
      "description": "Run open-source and fine-tuned models through Crusoe Cloud's OpenAI-compatible inference API, on serverless per-token endpoints or dedicated self-serve deployments.",
      "provider": "Crusoe",
      "categories": [
        "infrastructure"
      ],
      "accessMethods": [
        "api"
      ],
      "verified": true,
      "official": true,
      "featured": false,
      "sponsored": false,
      "authType": "apiKey",
      "why": null
    },
    {
      "name": "Prime Inference",
      "slug": "prime-inference",
      "url": "https://agentswitchboard.dev/agents/prime-inference",
      "iconUrl": "https://agentswitchboard.dev/icons/prime-inference.png?v=9d6473b3",
      "description": "Serves hosted open models and gateway access to third-party models through one OpenAI-compatible API, with rerank and classify endpoints and a CLI.",
      "provider": "Prime Intellect",
      "categories": [
        "infrastructure"
      ],
      "accessMethods": [
        "api",
        "cli"
      ],
      "verified": true,
      "official": false,
      "featured": false,
      "sponsored": false,
      "authType": "apiKey",
      "why": null
    },
    {
      "name": "SambaNova Cloud (SambaCloud)",
      "slug": "sambanova-cloud",
      "url": "https://agentswitchboard.dev/agents/sambanova-cloud",
      "iconUrl": "https://agentswitchboard.dev/icons/sambanova-cloud.png?v=f7eda7ee",
      "description": "Serves open models fast on SambaNova chips through an OpenAI-compatible inference API and Python SDK.",
      "provider": "SambaNova",
      "categories": [
        "infrastructure",
        "language"
      ],
      "accessMethods": [
        "api"
      ],
      "verified": true,
      "official": true,
      "featured": false,
      "sponsored": false,
      "authType": "apiKey",
      "why": null
    },
    {
      "name": "Cerebras Inference",
      "slug": "cerebras-inference",
      "url": "https://agentswitchboard.dev/agents/cerebras-inference",
      "iconUrl": "https://agentswitchboard.dev/icons/cerebras-inference.png?v=874fceca",
      "description": "Serves open-weight LLMs such as Qwen and GPT OSS through an OpenAI-compatible API on wafer-scale hardware, for developers who need very fast inference.",
      "provider": "Cerebras",
      "categories": [
        "language",
        "infrastructure"
      ],
      "accessMethods": [
        "api"
      ],
      "verified": true,
      "official": true,
      "featured": false,
      "sponsored": false,
      "authType": "apiKey",
      "why": null
    },
    {
      "name": "Nebius Token Factory",
      "slug": "nebius-token-factory",
      "url": "https://agentswitchboard.dev/agents/nebius-token-factory",
      "iconUrl": "https://agentswitchboard.dev/icons/nebius-token-factory.png?v=bd829350",
      "description": "Runs inference and fine-tuning on open models through Nebius's OpenAI-compatible API, with sandboxes for agents.",
      "provider": "Nebius",
      "categories": [
        "infrastructure",
        "language"
      ],
      "accessMethods": [
        "api"
      ],
      "verified": true,
      "official": true,
      "featured": false,
      "sponsored": false,
      "authType": "apiKey",
      "why": null
    },
    {
      "name": "Kimi API Platform",
      "slug": "kimi-api",
      "url": "https://agentswitchboard.dev/agents/kimi-api",
      "iconUrl": "https://agentswitchboard.dev/icons/kimi-api.png?v=e839cc9b",
      "description": "Serves Moonshot AI's Kimi models through OpenAI-compatible Chat Completions, Responses, and Anthropic-compatible Messages APIs, with tool calling, web search, and a 1M-token context.",
      "provider": "Moonshot AI",
      "categories": [
        "language",
        "infrastructure"
      ],
      "accessMethods": [
        "api"
      ],
      "verified": true,
      "official": true,
      "featured": false,
      "sponsored": false,
      "authType": "apiKey",
      "why": null
    },
    {
      "name": "Magnitude",
      "slug": "magnitude",
      "url": "https://agentswitchboard.dev/agents/magnitude",
      "iconUrl": "https://agentswitchboard.dev/icons/magnitude.png?v=a3fe107a",
      "description": "Runs open-weight models locally on kernels tuned to your hardware and connects them to coding agents through the magnitude CLI and an OpenAI- and Anthropic-compatible API.",
      "provider": "Magnitude",
      "categories": [
        "infrastructure"
      ],
      "accessMethods": [
        "cli",
        "api"
      ],
      "verified": true,
      "official": false,
      "featured": false,
      "sponsored": false,
      "authType": "none",
      "why": "Local Inference API: Serves OpenAI Chat Completions, OpenAI Responses, and Anthropic Messages routes on port 10100, as JSON or server-sent event…"
    },
    {
      "name": "OpenAI API",
      "slug": "openai-api",
      "url": "https://agentswitchboard.dev/agents/openai-api",
      "iconUrl": "https://agentswitchboard.dev/icons/openai-api.png?v=4dbfec46",
      "description": "Serves OpenAI models through the Responses and Chat Completions APIs, with function calling, file search, realtime voice, and remote MCP tools, for developers building model-backed apps.",
      "provider": "OpenAI",
      "categories": [
        "language",
        "infrastructure"
      ],
      "accessMethods": [
        "api",
        "cli"
      ],
      "verified": true,
      "official": true,
      "featured": false,
      "sponsored": false,
      "authType": "apiKey",
      "why": "Text Generation: Generates text, code, and structured JSON from prompts with OpenAI models through the Responses API."
    },
    {
      "name": "Groq",
      "slug": "groq",
      "url": "https://agentswitchboard.dev/agents/groq",
      "iconUrl": "https://agentswitchboard.dev/icons/groq.png?v=b1b9542c",
      "description": "Low-latency inference API running open models on custom LPU hardware, with an OpenAI-compatible interface and tool-calling support.",
      "provider": "Groq",
      "categories": [
        "language",
        "infrastructure"
      ],
      "accessMethods": [
        "api"
      ],
      "verified": true,
      "official": true,
      "featured": false,
      "sponsored": false,
      "authType": "apiKey",
      "why": null
    },
    {
      "name": "Fireworks AI",
      "slug": "fireworks-ai",
      "url": "https://agentswitchboard.dev/agents/fireworks-ai",
      "iconUrl": "https://agentswitchboard.dev/icons/fireworks-ai.png?v=485260d9",
      "description": "Production inference platform for open models with fast function calling, structured output, and fine-tuning behind an OpenAI-compatible API.",
      "provider": "Fireworks AI",
      "categories": [
        "language",
        "infrastructure"
      ],
      "accessMethods": [
        "api"
      ],
      "verified": true,
      "official": true,
      "featured": false,
      "sponsored": false,
      "authType": "apiKey",
      "why": null
    },
    {
      "name": "Jan",
      "slug": "jan",
      "url": "https://agentswitchboard.dev/agents/jan",
      "iconUrl": "https://agentswitchboard.dev/icons/jan.png?v=a8c129b1",
      "description": "Runs local LLMs 100% offline in a desktop app and connects cloud models, exposing a local OpenAI-compatible API. Privacy-first ChatGPT alternative by Menlo Research.",
      "provider": "Menlo Research",
      "categories": [
        "infrastructure"
      ],
      "accessMethods": [
        "api",
        "cli"
      ],
      "verified": true,
      "official": false,
      "featured": false,
      "sponsored": false,
      "authType": "none",
      "why": null
    },
    {
      "name": "Together AI",
      "slug": "together-ai",
      "url": "https://agentswitchboard.dev/agents/together-ai",
      "iconUrl": "https://agentswitchboard.dev/icons/together-ai.png?v=6c6dd240",
      "description": "Inference and fine-tuning platform for open models, offering serverless endpoints, dedicated GPU clusters, and an OpenAI-compatible API.",
      "provider": "Together AI",
      "categories": [
        "language",
        "infrastructure"
      ],
      "accessMethods": [
        "api"
      ],
      "verified": true,
      "official": true,
      "featured": false,
      "sponsored": false,
      "authType": "apiKey",
      "why": null
    },
    {
      "name": "vLLM",
      "slug": "vllm",
      "url": "https://agentswitchboard.dev/agents/vllm",
      "iconUrl": "https://agentswitchboard.dev/icons/vllm.png?v=77d62a59",
      "description": "High-throughput LLM inference server with OpenAI-compatible API. Serves Llama, Mistral, Qwen, and 30+ model families with PagedAttention. pip install vllm.",
      "provider": "vLLM Project",
      "categories": [
        "infrastructure"
      ],
      "accessMethods": [
        "api",
        "cli"
      ],
      "verified": true,
      "official": false,
      "featured": false,
      "sponsored": false,
      "authType": "none",
      "why": null
    },
    {
      "name": "Amazon Bedrock",
      "slug": "amazon-bedrock",
      "url": "https://agentswitchboard.dev/agents/amazon-bedrock",
      "iconUrl": "https://agentswitchboard.dev/icons/amazon-bedrock.png?v=fd9c599f",
      "description": "Serves 100+ foundation models from Anthropic, OpenAI, Amazon, DeepSeek, and others through one managed AWS API for teams building generative AI apps.",
      "provider": "Amazon Web Services",
      "categories": [
        "language",
        "infrastructure"
      ],
      "accessMethods": [
        "api"
      ],
      "verified": true,
      "official": true,
      "featured": false,
      "sponsored": false,
      "authType": "apiKey",
      "why": null
    },
    {
      "name": "Cloudflare Workers AI",
      "slug": "cloudflare-workers-ai",
      "url": "https://agentswitchboard.dev/agents/cloudflare-workers-ai",
      "iconUrl": "https://agentswitchboard.dev/icons/cloudflare-workers-ai.png?v=7c4f193b",
      "description": "Runs open-source models on serverless GPUs across Cloudflare's network from Workers, Pages, or the REST API, for text generation, embeddings, speech, and image tasks.",
      "provider": "Cloudflare",
      "categories": [
        "infrastructure",
        "language"
      ],
      "accessMethods": [
        "api"
      ],
      "verified": true,
      "official": true,
      "featured": false,
      "sponsored": false,
      "authType": "bearer",
      "why": null
    },
    {
      "name": "Colibri",
      "slug": "colibri",
      "url": "https://agentswitchboard.dev/agents/colibri",
      "iconUrl": "https://agentswitchboard.dev/icons/colibri.png?v=326bd367",
      "description": "Runs large Mixture-of-Experts models such as GLM, DeepSeek V4, and Kimi on consumer hardware by streaming routed experts from SSD, with an OpenAI-compatible local API.",
      "provider": "JustVugg",
      "categories": [
        "infrastructure"
      ],
      "accessMethods": [
        "cli",
        "api"
      ],
      "verified": true,
      "official": false,
      "featured": false,
      "sponsored": false,
      "authType": "none",
      "why": null
    },
    {
      "name": "ds4",
      "slug": "ds4",
      "url": "https://agentswitchboard.dev/agents/ds4",
      "iconUrl": "https://agentswitchboard.dev/icons/ds4.png?v=87563232",
      "description": "Runs DeepSeek V4 Flash, GLM 5.x, and Qwen models locally on Metal, CUDA, or ROCm with a native engine, a CLI, a coding agent, and an OpenAI-compatible server.",
      "provider": "Salvatore Sanfilippo (antirez)",
      "categories": [
        "infrastructure"
      ],
      "accessMethods": [
        "cli",
        "api"
      ],
      "verified": true,
      "official": false,
      "featured": false,
      "sponsored": false,
      "authType": "none",
      "why": null
    },
    {
      "name": "TensorFold",
      "slug": "tensorfold",
      "url": "https://agentswitchboard.dev/agents/tensorfold",
      "iconUrl": "https://agentswitchboard.dev/icons/tensorfold.png?v=be5b64e6",
      "description": "Serves local language models from a native Zig binary on Apple Silicon and NVIDIA GPUs with OpenAI- and Anthropic-compatible APIs, for developers running models on their own hardware.",
      "provider": "ashhart",
      "categories": [
        "infrastructure"
      ],
      "accessMethods": [
        "api",
        "cli"
      ],
      "verified": true,
      "official": false,
      "featured": false,
      "sponsored": false,
      "authType": "none",
      "why": null
    },
    {
      "name": "Xiaomi MiMo API Open Platform",
      "slug": "xiaomi-mimo-api",
      "url": "https://agentswitchboard.dev/agents/xiaomi-mimo-api",
      "iconUrl": "https://agentswitchboard.dev/icons/xiaomi-mimo-api.png?v=8de303f7",
      "description": "Calls Xiaomi's MiMo models through the MiMo API Open Platform's OpenAI-compatible endpoints.",
      "provider": "Xiaomi",
      "categories": [
        "language",
        "infrastructure"
      ],
      "accessMethods": [
        "api"
      ],
      "verified": true,
      "official": true,
      "featured": false,
      "sponsored": false,
      "authType": "apiKey",
      "why": null
    },
    {
      "name": "Baseten",
      "slug": "baseten",
      "url": "https://agentswitchboard.dev/agents/baseten",
      "iconUrl": "https://agentswitchboard.dev/icons/baseten.png?v=40ffb50e",
      "description": "Deploys models as production inference APIs with dedicated GPU deployments, Truss packaging, and OpenAI-compatible endpoints for agent workloads.",
      "provider": "Baseten",
      "categories": [
        "infrastructure"
      ],
      "accessMethods": [
        "api",
        "cli"
      ],
      "verified": true,
      "official": true,
      "featured": false,
      "sponsored": false,
      "authType": "apiKey",
      "why": "OpenAI-Compatible Inference: Serves open models behind an OpenAI-compatible API so existing agent clients switch backends by changing a base URL."
    },
    {
      "name": "Reka API",
      "slug": "reka-api",
      "url": "https://agentswitchboard.dev/agents/reka-api",
      "iconUrl": "https://agentswitchboard.dev/icons/reka-api.png?v=59ed822e",
      "description": "Serves Reka's multimodal models and selected open models through an OpenAI-compatible API, plus a Vision API and MCP server for video search and Q&A, for developers.",
      "provider": "Reka",
      "categories": [
        "language",
        "infrastructure",
        "content-media"
      ],
      "accessMethods": [
        "api",
        "mcp"
      ],
      "verified": true,
      "official": true,
      "featured": false,
      "sponsored": false,
      "authType": "apiKey",
      "why": "Multimodal Chat: Sends text, images and short videos to Reka and open models through OpenAI-compatible Chat Completions."
    },
    {
      "name": "LM Studio",
      "slug": "lm-studio",
      "url": "https://agentswitchboard.dev/agents/lm-studio",
      "iconUrl": "https://agentswitchboard.dev/icons/lm-studio.png?v=892d3112",
      "description": "Runs open LLMs locally with an OpenAI- and Anthropic-compatible server, the lms CLI, a headless daemon, and MCP support. New Bionic app adds an agent for coding, research, and docs. By Element Labs.",
      "provider": "Element Labs",
      "categories": [
        "infrastructure"
      ],
      "accessMethods": [
        "api",
        "cli",
        "mcp"
      ],
      "verified": true,
      "official": false,
      "featured": false,
      "sponsored": false,
      "authType": "none",
      "why": "Local Inference Server: Serves open models on localhost with OpenAI- and Anthropic-compatible endpoints for drop-in agent integration."
    },
    {
      "name": "OCI Generative AI",
      "slug": "oci-generative-ai",
      "url": "https://agentswitchboard.dev/agents/oci-generative-ai",
      "iconUrl": "https://agentswitchboard.dev/icons/oci-generative-ai.png?v=a84dec95",
      "description": "Provides Oracle Cloud's managed model inference, OpenAI-compatible Responses API, agent tools, vector stores and guardrails, for enterprises building governed AI apps and agents on OCI.",
      "provider": "Oracle",
      "categories": [
        "language",
        "infrastructure"
      ],
      "accessMethods": [
        "api",
        "cli"
      ],
      "verified": true,
      "official": true,
      "featured": false,
      "sponsored": false,
      "authType": "apiKey",
      "why": "Embeddings and Rerank: Runs chat, embedding and rerank tasks through the OCI-native inference API."
    },
    {
      "name": "LocalAI",
      "slug": "localai",
      "url": "https://agentswitchboard.dev/agents/localai",
      "iconUrl": "https://agentswitchboard.dev/icons/localai.png?v=f8a44eca",
      "description": "Serves local LLM, vision, image and voice inference behind an OpenAI-compatible API, letting agents run on consumer hardware with no GPU required.",
      "provider": "LocalAI",
      "categories": [
        "infrastructure"
      ],
      "accessMethods": [
        "api",
        "cli"
      ],
      "verified": true,
      "official": false,
      "featured": false,
      "sponsored": false,
      "authType": "none",
      "why": null
    },
    {
      "name": "oMLX",
      "slug": "omlx",
      "url": "https://agentswitchboard.dev/agents/omlx",
      "iconUrl": "https://agentswitchboard.dev/icons/omlx.png?v=3e00980e",
      "description": "Serves local LLM, vision, embedding and reranker models on Apple Silicon Macs through OpenAI- and Anthropic-compatible APIs, with continuous batching and an SSD-backed KV cache.",
      "provider": "Jun Kim",
      "categories": [
        "infrastructure"
      ],
      "accessMethods": [
        "cli",
        "api"
      ],
      "verified": true,
      "official": false,
      "featured": false,
      "sponsored": false,
      "authType": "apiKey",
      "why": null
    },
    {
      "name": "AI21 Studio",
      "slug": "ai21-studio",
      "url": "https://agentswitchboard.dev/agents/ai21-studio",
      "iconUrl": "https://agentswitchboard.dev/icons/ai21-studio.png?v=1828ea2c",
      "description": "Serves AI21's Jamba models and the Maestro agent-run API with file search, web search and remote MCP tools, for enterprises building grounded RAG agents.",
      "provider": "AI21 Labs",
      "categories": [
        "language",
        "infrastructure",
        "orchestration"
      ],
      "accessMethods": [
        "api"
      ],
      "verified": true,
      "official": true,
      "featured": false,
      "sponsored": false,
      "authType": "apiKey",
      "why": null
    },
    {
      "name": "Alibaba Cloud Model Studio",
      "slug": "alibaba-model-studio",
      "url": "https://agentswitchboard.dev/agents/alibaba-model-studio",
      "iconUrl": "https://agentswitchboard.dev/icons/alibaba-model-studio.png?v=3ce88d8f",
      "description": "Serves Qwen and third-party models through DashScope and OpenAI-compatible APIs on Alibaba Cloud, covering text, vision, image, speech, and embeddings, plus a RAG knowledge base.",
      "provider": "Alibaba Cloud",
      "categories": [
        "language",
        "infrastructure"
      ],
      "accessMethods": [
        "api"
      ],
      "verified": true,
      "official": true,
      "featured": false,
      "sponsored": false,
      "authType": "apiKey",
      "why": null
    },
    {
      "name": "Aperture by Tailscale",
      "slug": "tailscale-aperture",
      "url": "https://agentswitchboard.dev/agents/tailscale-aperture",
      "iconUrl": "https://agentswitchboard.dev/icons/tailscale-aperture.png?v=c24aedaf",
      "description": "Routes LLM and MCP traffic from coding agents through a gateway in your Tailscale network, with authentication by tailnet identity instead of shared API keys, plus spending limits and logs.",
      "provider": "Tailscale",
      "categories": [
        "infrastructure",
        "security"
      ],
      "accessMethods": [
        "api",
        "mcp",
        "cli"
      ],
      "verified": true,
      "official": false,
      "featured": false,
      "sponsored": false,
      "authType": "apiKey",
      "why": null
    },
    {
      "name": "apfel",
      "slug": "apfel",
      "url": "https://agentswitchboard.dev/agents/apfel",
      "iconUrl": "https://agentswitchboard.dev/icons/apfel.png?v=bce251ad",
      "description": "Runs Apple's on-device FoundationModels LLM from a CLI and a local OpenAI-compatible server, with MCP tool calling.",
      "provider": "Arthur Ficial",
      "categories": [
        "infrastructure"
      ],
      "accessMethods": [
        "cli",
        "api"
      ],
      "verified": true,
      "official": false,
      "featured": false,
      "sponsored": false,
      "authType": "none",
      "why": null
    },
    {
      "name": "Baidu Qianfan",
      "slug": "baidu-qianfan",
      "url": "https://agentswitchboard.dev/agents/baidu-qianfan",
      "iconUrl": "https://agentswitchboard.dev/icons/baidu-qianfan.png?v=7acc3db3",
      "description": "Serves Baidu's ERNIE and third-party models through an OpenAI-compatible API and hosts an agent development platform with tools, MCP and knowledge bases, for enterprises in China.",
      "provider": "Baidu",
      "categories": [
        "language",
        "infrastructure",
        "orchestration"
      ],
      "accessMethods": [
        "api"
      ],
      "verified": true,
      "official": true,
      "featured": false,
      "sponsored": false,
      "authType": "apiKey",
      "why": null
    },
    {
      "name": "Claude Developer Platform (Claude API)",
      "slug": "claude-developer-platform",
      "url": "https://agentswitchboard.dev/agents/claude-developer-platform",
      "iconUrl": "https://agentswitchboard.dev/icons/claude-developer-platform.png?v=f8f76b6a",
      "description": "Gives developers API access to Claude models through the Messages API, with streaming, tool use, batch processing, file uploads, and beta Managed Agents sessions.",
      "provider": "Anthropic",
      "categories": [
        "language",
        "infrastructure"
      ],
      "accessMethods": [
        "api"
      ],
      "verified": true,
      "official": true,
      "featured": false,
      "sponsored": false,
      "authType": "apiKey",
      "why": null
    },
    {
      "name": "ClawRouter",
      "slug": "clawrouter",
      "url": "https://agentswitchboard.dev/agents/clawrouter",
      "iconUrl": "https://agentswitchboard.dev/icons/clawrouter.png?v=524f918c",
      "description": "Routes agents' LLM calls through a local OpenAI-compatible proxy that pays per request in USDC over x402, with no accounts or API keys.",
      "provider": "BlockRun.ai",
      "categories": [
        "infrastructure",
        "commerce-payments"
      ],
      "accessMethods": [
        "cli",
        "api"
      ],
      "verified": true,
      "official": false,
      "featured": false,
      "sponsored": false,
      "authType": "none",
      "why": null
    },
    {
      "name": "Cloudflare Clef",
      "slug": "cloudflare-clef",
      "url": "https://agentswitchboard.dev/agents/cloudflare-clef",
      "iconUrl": "https://agentswitchboard.dev/icons/cloudflare-clef.png?v=7c4f193b",
      "description": "Serves Cloudflare's open-weight Clef and Clef-flash decision models on Workers AI, returning a probability for every option of typed questions about a text, JSON, or image state.",
      "provider": "Cloudflare",
      "categories": [
        "language",
        "infrastructure"
      ],
      "accessMethods": [
        "api"
      ],
      "verified": true,
      "official": true,
      "featured": false,
      "sponsored": false,
      "authType": "bearer",
      "why": null
    },
    {
      "name": "Cohere",
      "slug": "cohere",
      "url": "https://agentswitchboard.dev/agents/cohere",
      "iconUrl": "https://agentswitchboard.dev/icons/cohere.png?v=74fe4b3a",
      "description": "Provides Command models for chat, RAG, and tool use, plus Embed and Rerank models for search, via a REST API and Python, TypeScript, Java, and Go SDKs for enterprise developers.",
      "provider": "Cohere",
      "categories": [
        "language"
      ],
      "accessMethods": [
        "api"
      ],
      "verified": true,
      "official": true,
      "featured": false,
      "sponsored": false,
      "authType": "apiKey",
      "why": null
    },
    {
      "name": "DeepSeek API",
      "slug": "deepseek-api",
      "url": "https://agentswitchboard.dev/agents/deepseek-api",
      "iconUrl": "https://agentswitchboard.dev/icons/deepseek-api.png?v=c7b94b0a",
      "description": "Serves DeepSeek models through OpenAI- and Anthropic-compatible APIs with thinking mode, tool calls, JSON output, and context caching, for developers using existing SDKs.",
      "provider": "DeepSeek",
      "categories": [
        "language",
        "infrastructure"
      ],
      "accessMethods": [
        "api"
      ],
      "verified": true,
      "official": true,
      "featured": false,
      "sponsored": false,
      "authType": "apiKey",
      "why": null
    },
    {
      "name": "Echo",
      "slug": "echo",
      "url": "https://agentswitchboard.dev/agents/echo",
      "iconUrl": "https://agentswitchboard.dev/icons/echo.png?v=4fc20029",
      "description": "Routes each request across specialist open-weight models behind a single endpoint, allocating compute per task to cut inference cost while holding output quality.",
      "provider": "Tracer",
      "categories": [
        "infrastructure"
      ],
      "accessMethods": [
        "api"
      ],
      "verified": true,
      "official": false,
      "featured": false,
      "sponsored": false,
      "authType": "apiKey",
      "why": null
    },
    {
      "name": "Experiential",
      "slug": "experiential",
      "url": "https://agentswitchboard.dev/agents/experiential",
      "iconUrl": "https://agentswitchboard.dev/icons/experiential.png?v=fdbf6ec0",
      "description": "Routes requests across closed, open-source, local, and custom models behind one OpenRouter-compatible gateway, turning accumulated usage into a better-tuned model.",
      "provider": "Experiential Labs",
      "categories": [
        "infrastructure",
        "orchestration"
      ],
      "accessMethods": [
        "api",
        "cli"
      ],
      "verified": true,
      "official": false,
      "featured": false,
      "sponsored": false,
      "authType": "apiKey",
      "why": null
    },
    {
      "name": "Gemini API",
      "slug": "gemini-api",
      "url": "https://agentswitchboard.dev/agents/gemini-api",
      "iconUrl": "https://agentswitchboard.dev/icons/gemini-api.png?v=ebf423a0",
      "description": "Serves Google's Gemini models through the Gemini Developer API and Gen AI SDKs, with function calling, structured output, Live API voice, and image generation, for developers.",
      "provider": "Google",
      "categories": [
        "language",
        "infrastructure"
      ],
      "accessMethods": [
        "api"
      ],
      "verified": true,
      "official": true,
      "featured": false,
      "sponsored": false,
      "authType": "apiKey",
      "why": null
    },
    {
      "name": "Headroom",
      "slug": "headroom",
      "url": "https://agentswitchboard.dev/agents/headroom",
      "iconUrl": "https://agentswitchboard.dev/icons/headroom.png?v=995e9d4d",
      "description": "Compresses tool outputs, logs, RAG chunks, files, and history locally before they reach the model, as a Python or TypeScript library, a drop-in proxy, an agent wrapper, or an MCP server.",
      "provider": "Headroom Labs",
      "categories": [
        "infrastructure",
        "code-devtools"
      ],
      "accessMethods": [
        "api",
        "mcp",
        "cli"
      ],
      "verified": true,
      "official": false,
      "featured": false,
      "sponsored": false,
      "authType": "none",
      "why": null
    },
    {
      "name": "Hugging Face Inference Providers",
      "slug": "hugging-face-inference-providers",
      "url": "https://agentswitchboard.dev/agents/hugging-face-inference-providers",
      "iconUrl": "https://agentswitchboard.dev/icons/hugging-face-inference-providers.png?v=cdb9e63c",
      "description": "Routes open-model inference across partner providers such as Groq, Together, and Fireworks with one Hugging Face token, via an OpenAI-compatible API and Python and JS SDKs.",
      "provider": "Hugging Face",
      "categories": [
        "language",
        "infrastructure"
      ],
      "accessMethods": [
        "api"
      ],
      "verified": true,
      "official": true,
      "featured": false,
      "sponsored": false,
      "authType": "bearer",
      "why": null
    },
    {
      "name": "Kev",
      "slug": "kev",
      "url": "https://agentswitchboard.dev/agents/kev",
      "iconUrl": "https://agentswitchboard.dev/icons/kev.png?v=2c459eec",
      "description": "Answers yes/no, multiple-choice, and rating questions about text with calibrated probabilities, for developers self-hosting a Jev-compatible decision API.",
      "provider": "Jared Palmer",
      "categories": [
        "language",
        "infrastructure"
      ],
      "accessMethods": [
        "api"
      ],
      "verified": true,
      "official": false,
      "featured": false,
      "sponsored": false,
      "authType": "none",
      "why": null
    },
    {
      "name": "Kong AI Gateway",
      "slug": "kong",
      "url": "https://agentswitchboard.dev/agents/kong",
      "iconUrl": "https://agentswitchboard.dev/icons/kong.png?v=dc63fa08",
      "description": "Governs and meters LLM and agent traffic through a high-performance API gateway, with an MCP gateway that turns existing REST APIs into agent-callable tools.",
      "provider": "Kong Inc.",
      "categories": [
        "infrastructure",
        "orchestration",
        "security"
      ],
      "accessMethods": [
        "api",
        "cli",
        "mcp"
      ],
      "verified": true,
      "official": true,
      "featured": true,
      "sponsored": false,
      "authType": "apiKey",
      "why": null
    },
    {
      "name": "Meta Model API",
      "slug": "meta-model-api",
      "url": "https://agentswitchboard.dev/agents/meta-model-api",
      "iconUrl": "https://agentswitchboard.dev/icons/meta-model-api.png?v=8bdadce3",
      "description": "Serves Meta's Muse Spark, Muse Image, Muse Voice Transcribe, and SAM models through Responses, Chat Completions, and Messages-compatible APIs for app and agent developers.",
      "provider": "Meta",
      "categories": [
        "language",
        "infrastructure"
      ],
      "accessMethods": [
        "api"
      ],
      "verified": true,
      "official": true,
      "featured": false,
      "sponsored": false,
      "authType": "apiKey",
      "why": null
    },
    {
      "name": "MiniMax API Platform",
      "slug": "minimax-api",
      "url": "https://agentswitchboard.dev/agents/minimax-api",
      "iconUrl": "https://agentswitchboard.dev/icons/minimax-api.png?v=07c1b80c",
      "description": "Serves MiniMax's M-series coding and agent models through Anthropic- and OpenAI-compatible APIs, plus video, speech, image and music generation, for developers building agents.",
      "provider": "MiniMax AI",
      "categories": [
        "language",
        "infrastructure"
      ],
      "accessMethods": [
        "api"
      ],
      "verified": true,
      "official": true,
      "featured": false,
      "sponsored": false,
      "authType": "apiKey",
      "why": null
    },
    {
      "name": "Mistral AI",
      "slug": "mistral-ai",
      "url": "https://agentswitchboard.dev/agents/mistral-ai",
      "iconUrl": "https://agentswitchboard.dev/icons/mistral-ai.png?v=2aba084c",
      "description": "Serves open-weight and frontier models through La Plateforme, with an Agents API adding persistent agents, built-in connectors, and MCP client support.",
      "provider": "Mistral AI",
      "categories": [
        "language",
        "orchestration"
      ],
      "accessMethods": [
        "api"
      ],
      "verified": true,
      "official": true,
      "featured": false,
      "sponsored": false,
      "authType": "apiKey",
      "why": null
    },
    {
      "name": "Ollama",
      "slug": "ollama",
      "url": "https://agentswitchboard.dev/agents/ollama",
      "iconUrl": "https://agentswitchboard.dev/icons/ollama.png?v=a6bb20cf",
      "description": "Run LLMs locally — Llama, Mistral, Gemma, DeepSeek, and 100+ models via CLI and REST API. OpenAI-compatible endpoint at localhost:11434 for direct agent integration.",
      "provider": "Ollama",
      "categories": [
        "infrastructure"
      ],
      "accessMethods": [
        "api",
        "cli"
      ],
      "verified": true,
      "official": false,
      "featured": false,
      "sponsored": false,
      "authType": "none",
      "why": null
    },
    {
      "name": "OpenRouter",
      "slug": "openrouter",
      "url": "https://agentswitchboard.dev/agents/openrouter",
      "iconUrl": "https://agentswitchboard.dev/icons/openrouter.png?v=056efdc2",
      "description": "Unified API gateway routing requests across hundreds of language models from many providers, with automatic fallback, cost routing, and one OpenAI-compatible interface.",
      "provider": "OpenRouter",
      "categories": [
        "language",
        "infrastructure"
      ],
      "accessMethods": [
        "api"
      ],
      "verified": true,
      "official": true,
      "featured": false,
      "sponsored": false,
      "authType": "apiKey",
      "why": null
    },
    {
      "name": "Pi",
      "slug": "pi-agent-toolkit",
      "url": "https://agentswitchboard.dev/agents/pi-agent-toolkit",
      "iconUrl": "https://agentswitchboard.dev/icons/pi-agent-toolkit.png?v=0113027f",
      "description": "Build coding agents on a minimal harness — a unified LLM API across providers, a compact agent loop, a terminal UI, and MCP support, from Armin Ronacher and Mario Zechner.",
      "provider": "Earendil Works",
      "categories": [
        "code-devtools",
        "orchestration"
      ],
      "accessMethods": [
        "cli",
        "api",
        "mcp"
      ],
      "verified": true,
      "official": false,
      "featured": true,
      "sponsored": false,
      "authType": "apiKey",
      "why": "Unified LLM API: Presents one interface across model providers so an agent loop swaps models without changing its calling code."
    },
    {
      "name": "Replicate",
      "slug": "replicate",
      "url": "https://agentswitchboard.dev/agents/replicate",
      "iconUrl": "https://agentswitchboard.dev/icons/replicate.png?v=a059c34b",
      "description": "Run open-source AI models in the cloud — image generation, language models, video, audio via simple API calls.",
      "provider": "Replicate",
      "categories": [
        "content-media",
        "infrastructure"
      ],
      "accessMethods": [
        "api",
        "cli"
      ],
      "verified": true,
      "official": true,
      "featured": false,
      "sponsored": false,
      "authType": "apiKey",
      "why": null
    },
    {
      "name": "Tencent Hunyuan",
      "slug": "tencent-hunyuan",
      "url": "https://agentswitchboard.dev/agents/tencent-hunyuan",
      "iconUrl": "https://agentswitchboard.dev/icons/tencent-hunyuan.png?v=a54ea957",
      "description": "Calls Tencent's Hunyuan models on Tencent Cloud through OpenAI-compatible endpoints and Tencent Cloud SDKs.",
      "provider": "Tencent",
      "categories": [
        "language",
        "infrastructure"
      ],
      "accessMethods": [
        "api"
      ],
      "verified": true,
      "official": true,
      "featured": false,
      "sponsored": false,
      "authType": "apiKey",
      "why": null
    },
    {
      "name": "Volcengine Ark",
      "slug": "volcengine-ark",
      "url": "https://agentswitchboard.dev/agents/volcengine-ark",
      "iconUrl": "https://agentswitchboard.dev/icons/volcengine-ark.png?v=9d4d9e1b",
      "description": "Serves ByteDance's Doubao models and third-party models through Volcengine's API, with text, vision, image and video generation plus tool calling, for developers building on Volcengine.",
      "provider": "Volcengine (ByteDance)",
      "categories": [
        "language",
        "infrastructure"
      ],
      "accessMethods": [
        "api"
      ],
      "verified": true,
      "official": true,
      "featured": false,
      "sponsored": false,
      "authType": "apiKey",
      "why": null
    },
    {
      "name": "xAI API",
      "slug": "xai-api",
      "url": "https://agentswitchboard.dev/agents/xai-api",
      "iconUrl": "https://agentswitchboard.dev/icons/xai-api.png?v=7e71e335",
      "description": "Serves xAI's Grok models through an OpenAI-compatible Responses API with function calling, web search, voice, and image generation, for developers building on Grok.",
      "provider": "xAI",
      "categories": [
        "language",
        "infrastructure"
      ],
      "accessMethods": [
        "api"
      ],
      "verified": true,
      "official": true,
      "featured": false,
      "sponsored": false,
      "authType": "apiKey",
      "why": null
    },
    {
      "name": "Z.ai API Platform",
      "slug": "zai-api",
      "url": "https://agentswitchboard.dev/agents/zai-api",
      "iconUrl": "https://agentswitchboard.dev/icons/zai-api.png?v=295ec6d5",
      "description": "Serves Z.ai's GLM models through an HTTP API and official Python and Java SDKs, with function calling, web search, image and video generation, and document layout parsing.",
      "provider": "Z.ai",
      "categories": [
        "language",
        "infrastructure"
      ],
      "accessMethods": [
        "api"
      ],
      "verified": true,
      "official": true,
      "featured": false,
      "sponsored": false,
      "authType": "apiKey",
      "why": null
    }
  ],
  "relatedTasks": [
    "deploy-an-app",
    "transcribe-audio",
    "translate-text",
    "manage-cloud-infrastructure"
  ],
  "stacks": [
    "realtime-voice-agent"
  ]
}