LLM.howAsk LLM.how
Pricing reviewed September 7, 2026Review dates and sources appear on each record.

Access map

Every model has an address. Often several.

Direct APIs, hyperscale clouds, multi-model gateways, local runtimes, consumer apps, and creative suites—what each platform carries and when to use it.

Six access patterns

Start with the operating model.

Direct API

Newest first-party features, clearest model lifecycle, and one vendor relationship.

Cloud catalog

Enterprise identity, regions, procurement, networking, governance, and many model makers.

Model gateway

One interface for comparison, routing, fallback, and open-model access.

Local runtime

Private files, offline operation, fixed hardware cost, and full control over weights.

Consumer app

The fastest way to test whether a model helps before you build anything.

Creative platform

Purpose-built editing, asset management, timelines, voices, and media workflows.

Platform directory

Direct API

BytePlus LAS

Developer account and API billing

Seedance 2.5 video generation and editing.

Best for
Seedance 2.5 video generation and editing.
Watch for
Confirm current account availability, pricing and usage rights.
Family access references · verify exact variants

Platform overview reviewed 2026-09-08. Confirm current models and account access with the provider.

Official catalog or docs

Cartesia API

Developer account and API billing

Sonic speech synthesis.

Best for
Sonic speech synthesis.
Watch for
Confirm current account availability, pricing and usage rights.
Family access references · verify exact variants

Platform overview reviewed 2026-09-08. Confirm current models and account access with the provider.

Official catalog or docs

ElevenLabs API

Developer account and API billing

Speech synthesis, transcription and audio generation.

Best for
Speech synthesis, transcription and audio generation.
Watch for
Confirm current account availability, pricing and usage rights.
Family access references · verify exact variants

Platform overview reviewed 2026-09-08. Confirm current models and account access with the provider.

Official catalog or docs

OpenAI Platform

API account and project billing

OpenAI GPT, reasoning, realtime, speech, image, video, embedding, and moderation models.

Best for
First-party OpenAI features and the Responses API.
Watch for
Model and tool pricing are separate; pin production model IDs.
Family access references · verify exact variants

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

Claude Platform

API account and credits

Current Claude models, tools, prompt caching, batches, files, and model discovery.

Best for
First-party Claude features and stable model IDs.
Watch for
Preview headers and cloud availability can differ.
Family access references · verify exact variants

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

Google AI Studio / Gemini API

Google project and API key

Gemini, live audio, image, video, music, embeddings, and managed agent models.

Best for
Fast prototyping and direct Gemini developer access.
Watch for
Stable, preview, latest, and experimental IDs have different lifecycle guarantees.
Family access references · verify exact variants

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

xAI Console

Developer account and credits

Grok text/vision, search, tools, image, and video models.

Best for
First-party Grok API features and current-context tools.
Watch for
Long-context pricing and media pricing use separate rules.
Family access references · verify exact variants

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

Mistral Studio

Mistral account and billing

Mistral general, code, speech, OCR, safety, embedding, and specialist models.

Best for
First-party Mistral APIs and European deployment options.
Watch for
Licenses vary from Apache 2.0 to commercial or noncommercial terms.
Family access references · verify exact variants

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

Cohere Platform

Developer account

Command, Aya, Embed, Rerank, Transcribe, and private Model Vault deployment.

Best for
Enterprise RAG, retrieval, multilingual work, and private deployment.
Watch for
Check endpoint and SDK compatibility for the newest Command variants.
Family access references · verify exact variants

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

DeepSeek

Chat account or API credits

DeepSeek V4 Pro and Flash through OpenAI- and Anthropic-compatible interfaces.

Best for
Low-cost long-context reasoning and coding.
Watch for
Confirm data policy and use current IDs; legacy aliases can retire.
Family access references · verify exact variants

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

Alibaba Model Studio / Qwen

Alibaba Cloud account

Hosted Qwen language, vision, image, audio, and video families.

Best for
First-party Qwen access and Alibaba Cloud deployment.
Watch for
Availability, names, and regions differ from open checkpoints.
Family access references · verify exact variants

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

Moonshot / Kimi Platform

Developer account

Kimi long-context, general, vision, and coding models.

Best for
First-party Kimi APIs and long-context agents.
Watch for
Consumer and developer availability can differ by region.
Family access references · verify exact variants

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

Z.ai Platform

Developer account

GLM language, coding, vision, image, video, and agent models.

Best for
First-party GLM access and multilingual agents.
Watch for
Confirm the exact model/version and regional endpoint.
Family access references · verify exact variants

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

MiniMax Platform

Developer account

MiniMax language, speech, music, image, and Hailuo video models.

Best for
One vendor spanning agents and creative media.
Watch for
Different endpoints and billing units apply by modality.
Family access references · verify exact variants

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

Black Forest Labs API

Developer account

FLUX image generation and editing models.

Best for
First-party FLUX features and current premium models.
Watch for
Downloadable and hosted model licenses differ.
Family access references · verify exact variants

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

Stability AI Platform

Developer account

Stable image and creative service endpoints plus links to downloadable releases.

Best for
Hosted Stability models and editing services.
Watch for
Licensing and availability differ by model family.
Family access references · verify exact variants

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

AI21 Studio

Developer account

Jamba language models and AI21 document-oriented APIs.

Best for
Long-context enterprise generation and RAG.
Watch for
Check which Jamba versions remain active before production use.
Family access references · verify exact variants

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

Liquid AI Platform

Developer account

Liquid Foundation Model APIs and deployment guidance.

Best for
Efficient alternative architectures and compact deployment.
Watch for
Availability varies between hosted and downloadable releases.
Family access references · verify exact variants

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

Platform directory

Cloud catalog

Claude Platform on AWS

AWS account and platform credentials

Documented Fable 5 and Fable 5.1 access through Anthropic’s AWS platform.

Best for
AWS procurement with the Claude API interface.
Watch for
Distinct from Amazon Bedrock. Confirm region and supported features; Mythos availability is not established here.

This platform is tracked as an access path; specific family mapping is intentionally handled by its live official catalog.

Platform overview reviewed 2026-09-07. Confirm current models and account access with the provider.

Official catalog or docs

Amazon Bedrock

AWS account, model access, and region

Amazon Nova plus Anthropic, Meta, Cohere, AI21, Stability AI, and other partner models.

Best for
AWS governance, private networking, agents, guardrails, and multi-model procurement.
Watch for
Model and feature availability is regional and changes by provider.
Family access references · verify exact variantsPlus 1 more families in the catalog.

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

Microsoft Foundry

Azure subscription and region

OpenAI plus selected Anthropic, Meta, Mistral, DeepSeek, Cohere, xAI, NVIDIA, and Hugging Face models.

Best for
Azure identity, governance, evaluation, and enterprise deployment.
Watch for
Foundry has multiple catalog and deployment types; regions differ.
Family access references · verify exact variantsPlus 7 more families in the catalog.

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

Google Vertex AI Model Garden

Google Cloud project and region

Google models plus partner and open models including Anthropic and Meta.

Best for
Google Cloud governance, tuning, evaluation, and managed endpoints.
Watch for
Console catalog, managed APIs, and self-deployed models have different terms.
Family access references · verify exact variantsPlus 1 more families in the catalog.

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

Oracle Cloud Generative AI

OCI tenancy and region

Oracle-hosted Cohere and Meta model families plus dedicated AI clusters.

Best for
OCI-native enterprise and database-adjacent workloads.
Watch for
Check regional availability and dedicated-cluster commitments.
Family access references · verify exact variants

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

NVIDIA NIM / Build

Hosted trial or NVIDIA AI Enterprise

Certified containers and hosted APIs for NVIDIA, OpenAI open weights, Meta, Qwen, DeepSeek, GLM, and specialist models.

Best for
Portable NVIDIA-optimized serving from DGX Spark to data center.
Watch for
Certified profiles are model-, precision-, and GPU-specific.
Family access references · verify exact variants

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

Cloudflare Workers AI

Cloudflare account

Text, image, speech, embedding, reranking, classification, translation, and safety models at the edge.

Best for
AI features close to Cloudflare Workers and existing edge applications.
Watch for
Model limits and Workers integration constraints differ from self-hosting.
Family access references · verify exact variants

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

Platform directory

Model gateway

OpenRouter

Account, credits, optional BYOK

More than 400 text, image, audio, video, embedding, speech, and rerank models across many providers.

Best for
Comparing models and routing across providers through one interface.
Watch for
Choose provider and data-retention policies deliberately; capabilities vary by route.
Family access references · verify exact variantsPlus 7 more families in the catalog.

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

Together AI

Account and credits

Serverless and dedicated language, image, video, audio, embedding, rerank, and moderation models.

Best for
Current open-model APIs, dedicated endpoints, fine-tuning, and broad media coverage.
Watch for
Serverless and dedicated catalogs differ; watch quantization and lifecycle.
Family access references · verify exact variantsPlus 8 more families in the catalog.

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

Fireworks AI

Account and credits

Serverless and dedicated open models, fine-tunes, embeddings, rerankers, audio, and multimodal endpoints.

Best for
Fast open-model inference and custom deployment.
Watch for
Not every public checkpoint is serverless; deployment choice affects cost.

This platform is tracked as an access path; specific family mapping is intentionally handled by its live official catalog.

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

Replicate

Account and usage billing

Community and official language, image, video, audio, 3D, and utility models packaged as callable versions.

Best for
Trying creative and research models without managing GPUs.
Watch for
Models are published by different owners; inspect version, license, and hardware pricing.
Family access references · verify exact variants

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

Hugging Face Hub / Inference

Free browsing; hosted or provider billing for inference

The broadest public repository of model weights, cards, datasets, Spaces, and hosted inference providers.

Best for
Finding official weights, licenses, model cards, and deployment formats.
Watch for
A model page is not a guarantee that your hardware or chosen provider can run it.
Family access references · verify exact variantsPlus 12 more families in the catalog.

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

GroqCloud

Developer account

A curated high-speed catalog including Llama, GPT-OSS, Qwen, MiniMax, Whisper, and Groq Compound systems.

Best for
Very fast text inference, transcription, and tool-equipped open-model systems.
Watch for
The catalog is intentionally narrower and preview models can disappear quickly.
Family access references · verify exact variants

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

Cerebras Inference

Developer account

A curated catalog of open language and coding models on Cerebras hardware.

Best for
Extremely high token throughput for supported models.
Watch for
Model selection is smaller than general aggregators.
Family access references · verify exact variants

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

Platform directory

Local runtime

Ollama

Install locally

A curated library of quantized language, vision, embedding, and code models with a simple local API.

Best for
The fastest route to a useful local model service.
Watch for
Library tags hide quantization choices; inspect memory before pulling.
Family access references · verify exact variants

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

LM Studio

Desktop app

Discover, download, compare, and serve compatible local models, especially GGUF and MLX formats.

Best for
Visual local model exploration and an OpenAI-compatible local server.
Watch for
Select a downloaded local model for offline processing. Optional cloud features have a separate data boundary. Memory and architecture support depend on the model file.
Family access references · verify exact variants

Platform overview reviewed 2026-09-07. Confirm current models and account access with the provider.

Official catalog or docs

llama.cpp

Open-source build or packaged app

Portable GGUF inference across CPU, NVIDIA, Apple Silicon, AMD, Vulkan, and other backends.

Best for
Maximum portability and control over quantized local text/vision models.
Watch for
Compatibility depends on timely architecture support and model conversion.
Family access references · verify exact variants

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

vLLM

Open-source Linux server

High-throughput serving for many Hugging Face language and multimodal architectures.

Best for
Production-style self-hosted APIs and multi-user GPU servers.
Watch for
Supported architectures, quantizations, and GPU features vary by release.
Family access references · verify exact variantsPlus 2 more families in the catalog.

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

MLX / MLX Community

Apple Silicon and open tools

Apple-optimized model inference and community-converted language, vision, speech, and image models.

Best for
Large local models on unified-memory Macs.
Watch for
CUDA-first creative and training tools may not have equivalent MLX paths.
Family access references · verify exact variants

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

ComfyUI

Desktop, manual install, or cloud

Node-based workflows for open image, video, audio, 3D, and multimodal models.

Best for
Saved, repeatable creative pipelines with explicit model and processing stages.
Watch for
Community nodes are executable code; verify publishers and pin versions.
Family access references · verify exact variants

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

Platform directory

Consumer app

Hailuo

Product account; plan and rollout limits apply

MiniMax H3 hosted video creation.

Best for
MiniMax H3 hosted video creation.
Watch for
Confirm current account availability, pricing and usage rights.
Family access references · verify exact variants

Platform overview reviewed 2026-09-08. Confirm current models and account access with the provider.

Official catalog or docs

Doubao Pro

Product account; plan and rollout limits apply

Seedance 2.5 consumer rollout.

Best for
Seedance 2.5 consumer rollout.
Watch for
Confirm current account availability, pricing and usage rights.
Family access references · verify exact variants

Platform overview reviewed 2026-09-08. Confirm current models and account access with the provider.

Official catalog or docs

Cartesia Playground

Product account; plan and rollout limits apply

Try Sonic voices and speech generation.

Best for
Try Sonic voices and speech generation.
Watch for
Confirm current account availability, pricing and usage rights.
Family access references · verify exact variants

Platform overview reviewed 2026-09-08. Confirm current models and account access with the provider.

Official catalog or docs

ChatGPT

Free and paid plans

OpenAI chat, research, image, voice, agent, and file workflows; product availability differs from the API.

Best for
Trying OpenAI capabilities before building.
Watch for
A ChatGPT plan does not include API usage.
Family access references · verify exact variants

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

Claude

Free and paid plans

Claude chat, artifacts, projects, research, integrations, and coding products.

Best for
Interactive writing, analysis, research, and coding.
Watch for
Plan limits and model selection differ from API billing.
Family access references · verify exact variants

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

Gemini

Google account and optional paid plan

Gemini chat, research, connected Google services, creative media, and live experiences.

Best for
Using Gemini with Google’s consumer ecosystem.
Watch for
Consumer access does not map one-to-one to API models.
Family access references · verify exact variants

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

Grok

Free and SuperGrok plans

Grok chat, web/X context, voice, files, image, video, and connectors.

Best for
Interactive current-context and creative work.
Watch for
Shared weekly allowances and API billing are separate.
Family access references · verify exact variants

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

Le Chat

Free and paid plans

Mistral chat, search, files, images, and productivity workflows.

Best for
Trying Mistral without writing code.
Watch for
Product model names and developer model IDs may differ.
Family access references · verify exact variants

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

Meta AI

Web and Meta apps

Consumer assistant experiences powered by Meta’s Llama ecosystem across web, WhatsApp, Messenger, and Instagram.

Best for
Trying Meta’s consumer AI experience.
Watch for
The product does not expose every downloadable Llama checkpoint.
Family access references · verify exact variants

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

Platform directory

Creative platform

Jimeng AI

Product account; plan and rollout limits apply

Seedance video creation tools.

Best for
Seedance video creation tools.
Watch for
Confirm current account availability, pricing and usage rights.
Family access references · verify exact variants

Platform overview reviewed 2026-09-08. Confirm current models and account access with the provider.

Official catalog or docs

Runway

Web plan or API credits

Runway video, image, editing, transformation, and visual-effects models.

Best for
Cloud-first professional video iteration.
Watch for
Product features, API models, resolutions, and credit costs differ.
Family access references · verify exact variants

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

Luma Dream Machine

Web plan or API credits

Luma Ray video and Photon image model families.

Best for
Cinematic image-to-video, camera motion, and cloud creative work.
Watch for
Queue priority and resolution depend on the plan and endpoint.
Family access references · verify exact variants

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

Adobe Firefly

Adobe plan or Services API

Adobe image, video, audio, vector, and partner-model workflows inside Creative Cloud and APIs.

Best for
Design production, enterprise creative governance, and Adobe application integration.
Watch for
Partner models have their own terms and feature coverage.
Family access references · verify exact variants

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

Ideogram

Web plan or API

Ideogram image generation and editing models with strong text rendering.

Best for
Posters, brand concepts, typography, and design-heavy images.
Watch for
API and web features may launch on different schedules.
Family access references · verify exact variants

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

ElevenLabs

Web plan or API

Speech, transcription, dubbing, conversational agents, sound effects, and music.

Best for
Production voice and audio workflows.
Watch for
Voice cloning requires rights and consent; products use different billing units.
Family access references · verify exact variants

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

Suno

Consumer web and mobile plans

Full-song music generation, lyrics, styles, covers, extensions, and editing.

Best for
Rapid music ideas and complete song drafts.
Watch for
Review current plan rights before commercial use.
Family access references · verify exact variants

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

Udio

Consumer web plans

Music generation, remixing, extension, editing, and style-based workflows.

Best for
Song sketches, variations, and music experimentation.
Watch for
Review current rights, upload rules, and plan limits.
Family access references · verify exact variants

Platform overview reviewed 2026-08-09. Confirm current models and account access with the provider.

Official catalog or docs

Make the choice

Model first. Platform second. Evaluation always.

Use the catalog to shortlist a family. Then compare where it runs. Test the same representative inputs on the exact model ID, endpoint, region, and processing tier you intend to buy.

Browse the model catalog →