AI Model Pricing, All in One Place

AI safety company behind Claude — known for Constitutional AI, helpful harmless and honest responses, and industry-leading context windows up to 1M tokens.

Models6

LatestClaude Opus 4.6

Free tierFree tier on Claude

Free (Claude.ai) — $0

Pro — $20/mo

Google Gemini

Google's AI division powering Gemini, Imagen, Veo, and integrates deeply with Google Cloud, Workspace, and Search. Offers massive context windows and multimodal capabilities.

Models6

LatestGemini 2.5 Pro

Free tierGenerous free tier via Google AI Studio: 1M tokens/month for Gemini 2

Free (AI Studio) — $0

Pay-As-You-Go (AI Studio) — From $0.10/1M tokens

xAI (Grok)

xAI founded by Elon Musk, offering Grok — a real-time AI with web search built in, known for wit and personality. Grok 4.1 Fast is the most affordable flagship model.

API Free Credits — $0 (up to $175)

DeepSeek

Chinese AI lab offering highly capable open-weight models at extremely competitive prices. DeepSeek-V3.2 rivals GPT-4o-level performance at a fraction of the cost.

Models2

LatestDeepSeek Chat (V3.2)

Free tierFree tier available with rate limits

Free (API) — $0 (limited)

API Pay-Per-Use — From $0.27/1M tokens

Mistral AI

French AI lab with a strong open-source tradition. Offers state-of-the-art models including Codestral for code generation and Le Chat — a free ChatGPT alternative.

Models5

LatestMistral Large 3

Free tierLe Chat free tier: limited messages/day

Free (Le Chat) — $0

Le Chat Pro — €15/mo

Groq

LPU (Language Processing Unit) inference provider delivering the fastest inference speeds in the industry — up to 1000+ tokens/second. Linear pricing with no hidden costs.

Models6

LatestLlama 4 Scout (17Bx16E)

Free tierFree tier available with rate limits

Free Tier — $0

API Pay-Per-Use — From $0.05/1M tokens

Meta Llama

Meta's open-weight AI initiative — Llama models freely available for research and commercial use. The most popular open-source foundation model family.

Models2

LatestLlama 4 Scout

Free tierLlama models are open-weight and free to download and run

Open Weight (Download) — $0

API via Groq — From $0.11/1M tokens

Ollama

Run AI models locally or in the cloud. Ollama provides an easy-to-use runtime for open-source LLMs including Llama, Mistral, Phi, Gemma, and more.

Models6

LatestLlama 3.2 3B

Free tierOllama is free to download and run locally

Local (Free) — $0

Ollama Cloud — Free (beta)

Zhipu AI (GLM)

Chinese AI lab behind the GLM (General Language Model) series, including ChatGLM, GLM-4, andCharacterGLM. Known for strong bilingual Chinese-English performance.

Models3

LatestGLM-4

Free tierAPI free tier available with rate limits

Free Tier — $0

API Pro — From $0.1/1M tokens

MiniMax

Chinese AI company behind the Hailuo AI video generation platform and the MiniMax text API. Known for competitive pricing and strong video generation capabilities.

Models2

LatestMiniMax Text 01

Free tierHailuo AI has a free tier for video generation

Hailuo (Free) — $0

API Pay-Per-Use — From $0.01/1M tokens

Cohere

Canadian AI company specializing in enterprise LLM solutions. Powers the Command, Embed, and Rerank model families. Known for strong RAG and retrieval capabilities.

Models4

LatestCommand R7B

Free tierFree tier includes Command R7B and limited embeddings

Free Tier — $0

API Production — From $0.15/1M tokens

Stability AI

Leading open-source AI company behind Stable Diffusion image generation, Stable Video, and the Stability AI API. The reference for open-weight image and video generation.

Models2

LatestStable Diffusion 3 Medium

Free tierFree tier available for image generation (limited credits)

Free Tier — $0

API Pay-Per-Use — From $0.005/image

Perplexity (Sonar)

AI-powered search engine with built-in citation and real-time web access. The Sonar models are specifically designed for research and information retrieval tasks.

Models3

LatestSonar Small

Free tierFree tier on perplexity

Free (Web) — $0

Pro — $20/mo

Together AI

Managed inference platform for open-source models. Offers access to Llama, Mistral, Phi, Qwen, DeepSeek, and more — with a strong fine-tuning offering.

Models3

LatestDeepSeek V3 (Together)

Free tierFree tier with limited tokens per day

Free Tier — $0

Pay-As-You-Go — From $0.04/1M tokens

Fireworks AI

High-performance inference platform optimized for speed and throughput. Offers access to Mixtral, Llama, Qwen, and Fireworks' own Mixtral variants with industry-leading speed.

Models3

LatestMixtral 8x22B (Fireworks)

Free tierFree tier with limited tokens

Free Tier — $0

Pro — From $0.20/1M tokens