No results found. Try a different search term.
Top AI Models & Inference APIs on API.market
Browse our collection of high-quality AI Models & Inference APIs
Unified API gateway providing access to 78 AI models including GPT-4, Claude, Gemini, DeepSeek, Qwen, Mistral, and more. OpenAI-compatible.
Call Meta Llama 3.1 8B Instruct through OpenAI style chat completions and text completions endpoints. Perfect for chatbots and agents, personal and enterprise!
Get seamless access to all Claude models (Claude-4, Claude-3.5 and more) with our high-performance, cost-effective API.
🚀 Unlock cutting-edge AI with models from OpenAI, Claude, Google Gemini, and Meta Llama. Build transformative experiences with the best in AI! ✨
Fast, OpenAI-compatible Qwen3.8 27B inference for chat, coding, tools and structured outputs.
GPT-5 Nano is OpenAI fastest, cheapest version of GPT-5. It's great for summarization and classification tasks
Say Goodbye to Token Anxiety! Unlock the World's Best AI Models for the Cost of a Single Meal.
High-Speed, Low-Cost AI API with Extensive Language Model Support for Apps
GPT-4o (GPT-4 Omni) is the most advanced multimodal model (accepting text or image inputs and outputting text)
High Availability and Unlimited Calls for GPT 3.5 Turbo. We provide users with high-quality services
Which stocks does this story move? One embedding space for headlines, themes and listed stock symbols; rank symbols by cosine. OpenAI-compatible.
Extractive summarization for agents and pipelines: every output sentence is taken from the source, so a summary never invents a claim. OpenAI-compatible.
Rank documents by relevance to a query with Qwen3 Reranker 8B, guided by an optional task instruction.
Chat with Google's Gemma 4 26B A4B mixture-of-experts model, with optional images, video and thinking mode.
Send text, images, audio or video and get a reply as text and natural speech.
Turn each line of text into a 768-number embedding for semantic search, retrieval and clustering.
Ask questions about an image or a video and get a written answer from an open-source multimodal model.
Anthropic's Opus model for agentic coding and knowledge work, with adjustable thinking effort and optional image input.
Answer questions, classify and summarize short texts with Google's instruction-tuned FLAN-T5 XL.
Turn text into 768-dimensional embedding vectors for semantic search, clustering and duplicate detection with all-mpnet-base-v2.
Answer prompts, conversations and image questions with OpenAI's GPT-5.6 Terra, the balanced GPT-5.6 tier for everyday production work.
Answer prompts, conversations and image questions with OpenAI's GPT-5.6 Sol, the flagship GPT-5.6 tier for coding and multi-step reasoning.
Fast, low-cost text generation with OpenAI's GPT-5.6 Luna, with optional images, reasoning effort and verbosity.
Continue any text with Mistral AI's Mistral 7B v0.1 base model, with control over length, sampling and stop sequences.
Continue any text with Meta's Llama 3 8B base model, with control over length, sampling and stop sequences.
Continue any text with Meta's Llama 3 70B base model, with control over length, sampling and stop sequences.
Chat replies from Meta's Llama 2 7B Chat with your own system prompt, length and sampling controls.
Continue any text with Meta's Llama 2 7B base model, with control over length, sampling and stop sequences.
Continue any text with Meta's Llama 2 70B base model, with control over length, sampling and stop sequences.
Chat with Meta's 13-billion-parameter Llama 2 model, fine-tuned for dialogue, with your own system prompt.
Continue any text with the 13-billion-parameter base Llama 2 model from Meta.
Turn text into a 768-dimension vector for semantic search, clustering and similarity.
Turn text into a 1024-dimension vector for semantic search, retrieval and clustering.
Ask questions about an image or a video, or chat in plain text, with Qwen3-VL 8B Instruct.
Ask Google's open-weight Gemma 4 31B instruction model questions about text or an image and get a text answer.
Turn one text or a batch into dense vectors and sparse token weights for multilingual hybrid search.
Solve multi-step questions with IBM's 8B reasoning model and choose how much it reasons per request.
Chat, answer over documents and call tools with IBM's 8B long-context instruct model.
Chat, answer over documents and call tools with IBM's 32B long-context instruct model.
Answer, explain and write with IBM's 8B model fine-tuned for reasoning and instruction following, with a 128K context length.
Summarize, translate, answer questions and handle code tasks with IBM's open 8B instruction model.
Generate short answers, drafts and poems from a text prompt with Google's 2B instruction-tuned Gemma model.
Turn a list of texts into 3584-dimensional embedding vectors for search, clustering and RAG.
Send a prompt, an optional system prompt and an optional image to Anthropic's Claude Sonnet 4.6 and get its text answer.
Turn up to 100 texts per request into 5,120-dimension embedding vectors from the Llama 2 13B language model.
Chat with Alibaba's Qwen3.7-Plus, a multimodal model that reads text and images, with a 1 million token context window.
Generate answers, plans and long-form text with the open-weight Qwen3 235B instruction model.
Generate text from a prompt with gpt-oss-20b, OpenAI's 20b open-weight language model.
Send a text prompt to OpenAI's 120b open-weight model and get a text answer.
OpenAI's model for complex professional work, coding and multi-step reasoning, with adjustable reasoning effort and verbosity.
OpenAI's fastest and most cost-effective GPT-5 model, with text and image input.
Generate text from a prompt, a message list or images with OpenAI's GPT-5 Mini, with adjustable reasoning effort and verbosity.
Send prompts, chat messages and images to OpenAI's GPT-5 with adjustable reasoning effort and answer length.
Low-latency, low-cost text generation and image understanding with OpenAI's GPT-4o mini.
Generate text and ask about images with OpenAI's GPT-4o using a prompt or a full message list.
Send text and images to Kimi K2.6, Moonshot AI's open model built for long-horizon coding.
Ask Moonshot AI's Kimi K2 Thinking model a question and get a reasoned text answer.
Chat completions from Meta's 8 billion parameter Llama 3 model, with a system prompt and full sampling control.
Generate chat answers with Meta's 70 billion parameter Llama 3 Instruct model, with full control over sampling and the prompt template.
Chat with Meta's Llama 4 Scout Instruct: send a prompt and system prompt and get a text answer.
Generate chat answers and text with Meta's Llama 4 Maverick Instruct, a 17 billion parameter model with 128 experts.
Run IBM's open 8-billion-parameter Granite 3.3 model for chat, reasoning and instruction following with a 128K context.
Generate text from prompts with images, videos and audio using Google's Gemini 3.5 Flash, with optional thinking.
Send a text prompt with optional images, videos and audio to Google Gemini 3.1 Pro and get a text answer, with a choice of thinking level.
Send a text prompt with optional images, videos and audio to Google Gemini 3 Pro and get a reasoned text answer.
Send a text prompt with optional images, videos and audio to Google Gemini 3 Flash and get a fast text answer.
Generate text with DeepSeek V3.1 and switch on its reasoning mode per request.
Send a prompt, an optional system prompt and an optional image to Claude Sonnet 5 and get a text answer, with adjustable thinking effort.
Generate code, answers and documents with Claude Sonnet 4.5 from a text prompt and an optional image.
Turn one text or one image into a 768-number CLIP embedding for search, matching and clustering.
Turn texts in many languages into 1024-number embedding vectors for search, clustering and similarity.
Ask questions about an image and chat about it with the LLaVA v1.6 vision-language model (Vicuna-13B).
Turn an image into a 1024-dimension embedding for visual similarity search, deduplication and clustering.
Send prompts, conversations and images to OpenAI's GPT-4.1 and get text answers for complex tasks.
Generate text from a prompt, a system prompt and an optional image with Anthropic's Claude Haiku 4.5.
Send prompts, conversations and images to OpenAI's GPT-5.2, with adjustable reasoning effort and answer length.
Turn images into numeric embedding vectors with a large SigLIP 2 vision encoder.
Turn an image into a 1024-number SigLIP embedding vector for similarity search and clustering.
Embed text, images, video and audio into one shared vector space for search and similarity.
Ask a question about an image and get a text answer from Alibaba's Qwen-VL-Chat vision language model.
Coding, reasoning and image questions with OpenAI's GPT-5.1, with configurable reasoning effort and verbosity.
Embed text in many languages and images into one shared space, with adjustable vector length.
Embed text, images or audio into one shared 1024-dimensional vector space.
Send a prompt, an optional system prompt and an optional image to Anthropic's Claude Opus 4.7 and get its text answer.
Generate text and answer questions about images with OpenAI's fast, low-cost GPT-4.1 mini.
Ask questions about an image and chat about it with the LLaVA v1.6 vision-language model (Mistral-7B).
Send a text prompt with optional images and videos to Google's Gemini 2.5 Flash and get a text answer, with an adjustable thinking budget.
Send a prompt, an optional system prompt and an optional image to Anthropic's Claude Opus 4.6 and get its text answer.
Send a prompt, an optional system prompt and an optional image to Anthropic's Claude Fable 5 and get its text answer.
Ask a question about an image and get a short text answer, or generate a caption for it.
Caption an image, answer a question about it, or check how well a caption matches it.
Ask a small vision language model to describe an image or answer a question about it.
Turn several texts and image URLs into CLIP embeddings in one request, one 768-number vector per line.
OpenAI-compatible chat completions for qwen3-coder-30b, gpt-oss-20b, qwen/qwen3-30b-a3b and gemma-4-26b-a4b-it, with streaming and one flat price per call.
🚀 Unlock cutting-edge AI with models from OpenAI, Claude, Google Gemini, and Meta Llama. Build transformative experiences with the best in AI! ✨
Access GPT-5.6 Sol, Terra ,Luna, GPT-4.1, GPT-4o models directly with high availability and enjoy a 50% discount!
OpenAI-compatible chat completions with Qwen3-Next-80B-A3B-Instruct (tool-calling) on a dedicated RTX 4070 Ti SUPER
Access MiniMax M3 and DeepSeek V4 Flash through one fast, developer-friendly API gateway.
OpenAI-compatible API for GPT-5.6 Sol with High reasoning — drop-in for Cursor, Cline, Aider, and any OpenAI SDK.
A high-performance REST API for generating human-like text, answering complex questions, and powering your AI applications.
Fast OpenAI-compatible chat and reasoning API with streaming support and up to 512K-token context.
Production-grade DeepSeek chat completions on an OpenAI-compatible endpoint. deepseek-chat & deepseek-reasoner, streaming supported, built for developers.
Fast text translation and language detection powered by LibreTranslate.
Integrate a powerful OpenAI-compatible GPT-style API featuring predictable low-cost token billing and robust stable metering.
Seamlessly integrate with powerful LLMs, offering OpenAI-style completions and blazing performance from a remote GPU cluster.
⚡50% Discount | Direct and highly available API for all latest Gemini models: Gemini 3 Series, Gemini 2.5 Series, and more
Unify is your centralized platform for LLM endpoints.
Experience low-latency chat completions with GLM-4, Qwen-Turbo, and DeepSeek at unmatched performance and cost savings.
Rapidly improve AI output accuracy by leveraging the most complete LLM Hallucination Taxonomy, Benchmark Scores, and Detection Methods.
GPT-4.1 mini excels at instruction following and tool calling. It features a 1M token context window, and low latency without a reasoning step.
GPT-4.1 nano excels at instruction following and tool calling. It features a 1M token context window, and low latency without a reasoning step.
GPT-5 mini is a faster, more cost-efficient version of GPT-5. It's great for well-defined tasks and precise prompts.
Generate high-quality images from text or image references quickly with impressive performance and flat pricing.
Google's fastest Gemini model. Ultra-low latency, multimodal, China direct connection.
Generate photorealistic images from text or reference images using Google's Gemini 3 Pro Image Preview.
Iteratively create and edit images with text, boasting impressive speed and multi-turn editing.
ByteDance's ultra-low cost, high-concurrency model. Multimodal, China direct connection.
Access the latest DeepSeek V4 Flash model. Fast, affordable, China direct connection, OpenAI compatible.