Best LLM APIs for Agentic Workflows

llm · optimization · rag

LLM application framework with auto-optimization — auto-tune prompts and RAG pipelines like PyTorch

evaluation · benchmark · llm

Automated LLM-based evaluation framework for AI agent tasks and benchmarks

unified-api · multi-provider · python

Simple, unified Python interface to multiple LLM providers by Andrew Ng. Call OpenAI, Anthropic, Groq, Google, Mistral, and more through one consistent API. Makes switching providers trivial.

llm · nlp · enterprise

Jamba and Jurassic LLM models with enterprise NLP solutions

llm · european ai · sovereign ai

European sovereign AI company providing Luminous LLMs and the PharIA model for enterprise and government use.

llm · aws · multimodal

Amazon's frontier multimodal model family (Nova Micro/Lite/Pro/Premier) for text, vision, and video. Native to AWS Bedrock with best-in-class cost/performance ratio.

framework · anthropic · claude

Anthropic's official Python SDK for building Claude-powered agents with tool use and streaming.

scraping · automation · data

Web scraping and automation platform with 1500+ ready-made actors for AI data pipelines

observability · monitoring · llm

ML observability platform with full LLM and agent monitoring. Detect hallucinations, trace agent runs, and debug production AI.

voice · stt · audio

Speech-to-text and audio intelligence API — transcription, summarization, and speaker detection

evaluation · llm · automated

Quickly evaluate LLM outputs using model-graded, heuristic and statistical methods

rag · optimization · open-source

Automated RAG optimization tool that finds the best RAG pipeline configuration for your data automatically.

enterprise · azure · openai

Enterprise-grade OpenAI models on Azure with compliance, security, and private networking.

agents · task-management · openai

AI-powered task management system using OpenAI and vector databases

llm · chinese-ai · open-source

Chinese AI company offering large language models with strong Chinese language understanding and generation.

deployment · inference · cloud

ML model deployment platform — fast, scalable inference for any open-source model

serverless · gpu · python

Serverless Python cloud for AI workloads — run GPU functions, fine-tuning, and inference at scale

eval · testing · observability

Enterprise AI evaluation platform. Log, test, and evaluate LLM applications with dataset management and CI/CD integration.

search · api · privacy

Independent web search API — privacy-focused search data for AI apps, free from Big Tech

tts · voice · api

Ultra-low-latency TTS API for real-time voice AI — sub-100ms streaming speech synthesis

inference · hardware · speed

World's fastest AI inference — up to 2,000 tokens/sec with Cerebras wafer-scale chips

inference · speed · llm

World's fastest LLM inference. 2,000+ tokens/second for Llama 3.3 70B. Purpose-built Cerebras chips deliver 20x faster inference than GPU clouds. Free tier available.

gpu · serverless · inference

Serverless GPU platform for AI inference — deploy ML models with auto-scaling and cold starts < 1s

computer vision · llm · platform

End-to-end AI platform for building, deploying, and scaling computer vision and LLM-powered applications.

llm · developer · model

Anthropic's family of large language models — safe, helpful, and honest

llm · anthropic · claude

Constitutional AI model by Anthropic — safe, harmless, and helpful LLM.

llm · anthropic · reasoning

Anthropic's most intelligent model with extended thinking for complex reasoning and coding tasks.

framework · anthropic · sdk

Anthropic's official SDK for building agents with Claude. Provides tool use, multi-step reasoning, and computer use capabilities.

anthropic · api · foundation-model

Anthropic's Claude API — access Claude 3.5/3.7 Sonnet/Opus for building safe and capable AI agents

coding · terminal · agentic

Anthropic's agentic coding tool operating directly in your terminal. 80.9% SWE-bench score. Reads your codebase, writes and runs code, tests, and deploys — with full git integration.

llm · foundation-model · agentic

Anthropic's latest frontier model (April 2026). SWE-bench 87.6%, 1M token context, advanced vision.

edge · inference · serverless

Run AI models at the edge globally. Workers AI provides serverless GPU inference for 50+ models with zero cold starts.

coding-agent · cli · openai

OpenAI's open-source terminal-based coding agent that reads your codebase, writes code, runs tests, and fixes bugs autonomously.

llm · enterprise · open-weights

Cohere's flagship enterprise LLM. 111B parameters, 256K context, open-weights. Optimized for agentic tasks, tool use, and multilingual enterprise RAG with grounded citations.

llm · rag · enterprise

Enterprise-grade LLM optimized for RAG and tool use — Cohere's flagship model for production agents

embeddings · rag · search

Enterprise-grade multilingual embeddings API for semantic search, classification, and RAG pipelines.

enterprise · llm · embeddings

Enterprise NLP platform with Command, Embed, and Rerank models optimized for business applications

llm · cohere · enterprise

Cohere's most capable enterprise LLM with advanced RAG support and business-grade reliability.

evaluation · testing · deepeval

LLM evaluation and testing platform powering DeepEval with regression testing and A/B testing

crawling · scraping · open-source

Open-source, LLM-friendly web crawling library. Extracts structured data from any website optimized for AI workflows.

image-generation · multimodal · openai

OpenAI's most capable image generation model with precise prompt following and high visual fidelity.

image-gen · design · openai

ml-platform · data · llm

Unified data and AI platform for building, training, and deploying ML models and AI agents at scale. Includes MLflow, Unity Catalog, and DBRX open-source LLM.

evaluation · testing · llm

Unit testing framework for LLM apps with 14+ built-in metrics. Hallucination detection, RAG evaluation, works like Pytest.

inference · serverless · open-source-models

Cheap and fast serverless model inference — hundreds of open-source models via API

education · courses · llm

Andrew Ng's AI education platform. Short courses on LLMs, AI agents, RAG, fine-tuning, and more. Partnered with OpenAI, Anthropic, Google, AWS for up-to-date practitioner content.

agent · api · framework

DeepSeek-V3.2 with enhanced Agent capabilities, integrated reasoning, available on web, app, and API

llm · api · deepseek

DeepSeek's API service offering access to DeepSeek-V3 and DeepSeek-R1 reasoning models, known for strong coding and math performance.

llm · reasoning · open-source

DeepSeek's next-generation reasoning model with enhanced mathematical and scientific problem-solving.

prompt engineering · llm · open-source

Language model programming library that treats prompts as functions, with built-in versioning and visualization.

evaluation · monitoring · open-source

Open-source ML and LLM observability platform for evaluating, testing, and monitoring model quality in production.

search · api · semantic

Semantic search API for AI apps — searches the web by meaning, not keywords

llm · lg · korean

LG AI Research's bilingual Korean-English language model for enterprise and research applications.

inference · quantization · gpu

Optimized inference library for quantized LLMs on consumer GPUs — fast GPTQ/EXL2

inference · serverless · image-generation

Serverless inference platform for generative AI models. Fast FLUX, SDXL, video generation, and audio models via simple API. GPU-accelerated, pay-per-use, 100ms cold starts.

api · framework

OpenClaw quick-start guide covering core concepts, use cases, and first deployment

scraping · api · llm

Web scraping API for LLMs — turns any website into clean Markdown for AI ingestion

cloud · inference · low-latency

Ultra-low latency LLM inference platform. FireFunction supports tool calling, ideal for latency-sensitive AI agent apps.

agent-platform · llm · deploy

Platform for building and deploying AI agents with LLM backends

api

Agent wallet for API payments, send to any address, and receive from anyone

llm · foundation-model · multimodal

Google's flagship multimodal model (2026). Best-in-class on long-context reasoning and coding tasks.

llm · multimodal · api

Google's multimodal LLM API. Gemini 2.5 Pro/Flash with 1M+ context window, native audio/video/image understanding, code execution, and grounding with Google Search.

llm · open-source · google

Google's open-weight model family for on-device and research use. Available in 1B–27B sizes. Instruction-tuned variants for chat and agentic tasks. Apache 2.0 licensed.

testing · observability · llm

AI pipeline testing and observability platform for evaluating, monitoring, and improving LLM outputs in production.

function-calling · api · llm

UC Berkeley LLM trained for API calls — accurately selects and invokes APIs from 1600+ tools.

llm · openai · gpt-4

Advanced multimodal language model by OpenAI — industry-leading reasoning.

ai · llm · model

Advanced language model by OpenAI, works with OpenClaw

llm · openai · multimodal

OpenAI's omni model with native multimodal understanding of text, vision, and audio in real-time.

llm · openai · reasoning

OpenAI's flagship frontier model with enhanced reasoning and agentic capabilities.

llm · foundation-model · agentic

OpenAI's frontier agentic model released April 2026. Significant gains in coding, online research, and long-horizon task handling.

ai-coding · codebase-understanding · code-review

AI code assistant that understands your entire codebase to answer questions, write code, and review PRs with full repository context.

inference · llm · fast

Fastest LLM inference on custom hardware

observability · monitoring · cost-tracking

LLM observability platform for monitoring costs, latency, and quality of AI applications. One-line integration.

evaluation · benchmark · llm

Holistic Evaluation of Language Models by Stanford CRFM — comprehensive multi-metric LLM benchmarking framework.

inference · serverless · huggingface

Serverless inference API for 150k+ Hugging Face models — no infrastructure setup needed

evaluation · observability · llm

AI evaluation platform for automated testing, tracing, and continuous monitoring of LLM pipelines.

voice · emotion · api

Emotionally intelligent voice AI. EVI (Empathic Voice Interface) API detects emotional cues in speech and adapts responses accordingly. Ultra-low latency, natural prosody, developer API.

llm · chinese-ai · tencent

Tencent's large language model with multimodal capabilities, supporting text, image, and code generation.

llm · personal-ai · enterprise

Enterprise AI API provider — Pi personal AI and powerful foundation models for business applications.

structured-output · pydantic · python

Python library for structured LLM outputs — validates and retries with Pydantic models

embeddings · api · multimodal

Multimodal embedding and search API — state-of-the-art embeddings for text, image, and code

llm · chinese ai · long context

Kimi API: advanced code gen, chat, visual reasoning, multimodal — built for complex tasks

llm · chinese-ai · long-context

Moonshot AI's long-context LLM with 1M token window — strong Chinese and English reasoning.

local-llm · inference · open-source

Easy-to-use local LLM inference backend for running GGUF models with web UI and API compatibility

framework · llm · rag

Building applications with LLMs through composability

observability · tracing · llm

Hosted version of Langfuse — LLM observability, tracing, and evaluation platform with managed infrastructure

observability · tracing · llm

Open-source LLM observability tool for tracing, evaluating, and debugging AI agents and LLM applications.

cloud · deployment · llm

AI application cloud platform. Deploy Python functions as APIs with built-in LLM inference and task queues.

inference · open-source · performance

High-performance LLM inference framework optimized for speed and efficiency in production deployments.

search · api · realtime

Real-time web search API for AI agents. Returns structured, clean content from the live web without scraping complexity.

llm · gateway · api

Unified LLM API gateway — call 100+ LLMs with one OpenAI-compatible interface

llm · open-source · meta

Meta's latest open-weight model family. Llama 4 Scout (10M token context, MoE, runs on single H100) and Llama 4 Maverick (1M context, 400B total params). Free to use and modify.

local-llm · inference · open-source

High-performance LLM inference in pure C/C++. Runs quantized models (GGUF) on CPU and GPU with minimal dependencies. The engine behind Ollama, LM Studio, and Jan. 75K+ GitHub stars.

local-llm · inference · open-source

Efficient C/C++ implementation for running LLMs locally with CPU and GPU support. Enables running Llama, Mistral, and other models on consumer hardware.

parsing · rag · pdf

Advanced document parser for RAG — extracts structured data from PDFs, tables, and complex docs

framework · meta · open-source

Meta's open-source stack for building AI agent applications with Llama models. Standardizes inference, safety, and agentic APIs.

training · finetuning · llm

MosaicML's codebase for training, finetuning, and deploying LLMs — optimized for efficiency at scale.

prompt programming · llm · open-source

Programming language for LLMs that enables constrained decoding, scripted prompting, and efficient token generation.

local · inference · open-source

Free OpenAI-compatible local inference server — run LLMs, image gen, and audio models on your hardware

python · llm · open-source

Pythonic AI engineering toolkit by Prefect for building reliable natural language interfaces and AI functions.

evaluation · testing · llm

AI quality platform for testing and evaluating LLM and agent applications before production.

mcp · filesystem · open-source

Official MCP server for filesystem operations.

mcp · github · open-source

Official MCP server for GitHub API operations.

llm · small-model · reasoning

Microsoft's small language model (14B) that punches above its weight on complex reasoning tasks. Ideal for edge deployment and cost-sensitive agent workloads.

framework · python · type-safe

Type-safe Python library for building LLM applications. Decorator-based API for calling LLMs, extracting structured data, and building tools. Works with OpenAI, Anthropic, Gemini, Mistral, and more.

framework · mistral · api

Mistral's native agentic API for building stateful, multi-turn agent workflows with tool support.

embeddings · reranking · rag

State-of-the-art embedding and reranking models for semantic search and RAG pipelines.

local-llm · mobile · inference

Machine Learning Compilation framework for LLMs enabling native deployment across any hardware — iOS, Android, WebGPU, CUDA, and Metal.

mcp · protocol · standard

Open standard by Anthropic for connecting AI assistants to data sources and tools. The USB-C of AI integrations.

integrations · api · oauth

Open-source product integrations platform — OAuth, webhooks, and data sync for 300+ APIs in minutes.

routing · llm · cost-optimization

AI model router that automatically selects the best LLM for each query. Learns from your prompts to route between GPT-4o, Claude, Gemini, and others — cutting costs by 30-50% while maintaining quality.

api · image-gen · llm

100+ AI APIs for image, video, and LLM generation at competitive pricing

inference · gpu · llm

High-performance AI inference cloud for LLMs and image generation. Runs Llama, Mistral, SDXL, and fine-tuned models with industry-leading latency. OctoStack for private cloud deployment.

openai · sdk · multi-agent

OpenAI's official Python SDK for building multi-agent systems with handoffs, tools, and guardrails

api · stateful · rag

OpenAI's stateful agent API. Built-in thread management, file search (RAG), code interpreter, and function calling. Build AI assistants without managing context windows.

openai · gpt · cookbook

Example code and guides for OpenAI API

research-agent · openai · autonomous

OpenAI's autonomous research agent. Conducts 5-30 minute deep research sessions with citation-backed reports using o3 model.

voice · api · realtime

WebSocket-based API for low-latency speech-to-speech voice agents. Sub-300ms response time, parallel function calling, and natural interruption handling for conversational AI.

framework · openai · multi-agent

OpenAI's experimental lightweight multi-agent orchestration framework. Core primitives: Agents and Handoffs. Educational reference implementation for building multi-agent systems. Predecessor to the OpenAI Agents SDK.

evaluation · benchmark · llm

Open-source LLM evaluation framework supporting 100+ benchmarks across reasoning, knowledge, and coding.

observability · opentelemetry · llm

OpenTelemetry-based observability for LLMs and AI agents by Traceloop

fine-tuning · training · llm

Turn your LLM API calls into fine-tuned models — replace GPT-4 with cheaper custom models

llm-gateway · api · routing

Unified API for accessing multiple LLM providers

computer-use · web-automation · autonomous-agent

OpenAI's computer-use agent that navigates and completes tasks on the web autonomously

enterprise · oracle · cloud

Oracle's AI services integrated with Oracle Cloud, offering LLM APIs, vector search, and agent tools.

structured-output · json · llm

Structured text generation library — guarantees valid JSON, regex, and grammar outputs from LLMs

prompt engineering · evaluation · testing

LLM engineering platform for prompt versioning, testing, and evaluation — built for teams shipping AI features fast.

evaluation · testing · safety

Automated evaluation platform for LLM applications with hallucination detection and safety testing

search · open-source · self-hosted

Open-source AI-powered search engine. Self-hostable alternative to Perplexity AI using local LLMs and web search.

search · llm · real-time

Conversational search engine powered by AI

prompt management · observability · open-source

Open-source AI development toolkit — centralize prompt management, observe LLM usage, and troubleshoot AI in real-time.

llm · slm · microsoft

Microsoft's small language model family. Phi-4 (14B) rivals much larger models on reasoning; Phi-4-mini (3.8B) for edge deployment; Phi-4-multimodal integrates speech, vision, and text.

gateway · observability · llm

AI gateway with observability, prompt management and reliability for LLM apps

llm · workflow · microsoft

Microsoft Azure Prompt Flow — build, evaluate, and deploy high-quality LLM apps with visual DAG editor.

prompt-management · monitoring · observability

Prompt engineering and LLM monitoring platform — version control for prompts

llm · open-source · alibaba

Alibaba's open-source LLM series. Qwen3 (latest): 235B MoE, 32B dense, 0.6B–30B models. Top open-source benchmarks, 119 languages, thinking mode (chain-of-thought), Apache 2.0.

model serving · llm · open-source

Scalable model serving library built on Ray for deploying LLMs and ML models at production scale.

parsing · rag · pdf

Document parsing API for RAG — accurate extraction from PDFs, tables, and complex layouts

design · image-editing · background-removal

AI-powered background removal for images.

inference · models · api

Run and fine-tune open-source models

visual programming · llm · open-source

Open-source visual AI programming environment for building and debugging complex LLM-powered agent pipelines.

gpu · cloud · training

GPU cloud for AI workloads — cost-effective compute for training and running AI agents at scale

inference · hardware · enterprise

Ultra-fast AI inference platform with SambaNova custom RDU hardware

evaluation · testing · llm

LLM evaluation and testing platform — regression tests, red-teaming, and CI/CD for AI

scraping · llm · extraction

LLM-powered web scraping library. Define what data you want in natural language and let AI extract it from any webpage.

search · api · google

Google Search API for developers — structured search results for AI apps and agents

search-api · web-search · google

Fast, affordable Google Search API for AI agents. 2,500 free queries, then $0.30/1000. Returns structured JSON results including organic, images, news, and knowledge graph. Used by LangChain, CrewAI, and hundreds of agent frameworks.

inference · serving · llm

Fast LLM serving framework — structured generation language with 5x throughput gains

api · model

China's leading model aggregation platform with 200+ open-source and commercial models via unified API

coding-agent · open-source · scaffolding

Your personal junior developer AI agent that scaffolds entire apps from a product spec prompt

llm · upstage · korean

Upstage's top-performing enterprise LLM with strong multilingual reasoning and document processing.

video-generation · multimodal · openai

OpenAI's video generation model. Creates realistic videos up to 60 seconds from text prompts or images. Available in ChatGPT Plus and Pro.

integration · api · enterprise

Universal API platform for AI agents. 10,000+ actions across 200+ SaaS connectors — HRIS, CRM, ATS — enabling agents to act on enterprise software.

agent · framework · open-source

Open-source framework for building, deploying and managing AI agents with APIs, memory and tool use.

search · api · agent

Search API built for AI agents — fast, accurate web search with structured results

inference · nvidia · optimization

NVIDIA's open-source library for optimized LLM inference on NVIDIA GPUs. Quantization (FP8/INT4), in-flight batching, tensor parallelism, and 4x+ throughput vs standard PyTorch.

local-llm · webui · open-source

Popular Gradio-based web UI for running local LLMs supporting Llama, Mistral, Phi, and GGUF models

optimization · llm · research

Automatic differentiation via text — optimize LLM prompts and pipelines using textual gradients

inference · serving · llm

Hugging Face's production-ready server for serving LLMs — optimized inference for Llama, Mistral, and more.

inference · llm · open-source

Fast, affordable inference for open-source models

observability · opentelemetry · tracing

LLM observability via OpenTelemetry — open-source tracing and monitoring for AI applications

inference · nvidia · production

NVIDIA's production-grade model serving platform — multi-framework, GPU/CPU inference

voice · sms · api

Cloud communications platform. Programmable SMS, voice, WhatsApp, and email APIs. ConversationRelay for building AI voice agents with LLMs. Used by millions of developers globally.

ai · llm

Platform focused on LLM evaluation and benchmarking to improve generative AI performance.

text-to-sql · analytics · llm

Open-source AI SQL agent — ask questions in natural language, get accurate SQL queries automatically.

security · prompt-injection · llm

LLM prompt injection and jailbreak detection library for Python.

inference · serving · open-source

Fast and memory-efficient LLM serving engine. PagedAttention for high throughput, continuous batching, and OpenAI-compatible API server. 30K+ GitHub stars. The production standard for self-hosted LLMs.

speech · transcription · open-source

OpenAI's open-source automatic speech recognition (ASR) model, supports 99 languages with high accuracy.

llm · grok · api

xAI's Grok API — access Grok-2 and Grok-3 models for developers

agent · api · framework

Quality AI Agent skill sharing and evaluation platform

inference · serving · open-source

Open-source model serving platform — deploy LLMs, vision models, and embeddings locally or in cloud

integration · agent-tools · api

AI agent integration platform with 1,000+ pre-built tools. Deploy production agents with memory, sessions, and enterprise-grade reliability.

llm · chinese-ai · open-source

High-performance bilingual LLM by 01.AI (Kai-Fu Lee) with strong Chinese and English capabilities.

llm

Zhipu AI's open platform for GLM large language models with multimodal capabilities.