📊LLM API Pricing Guide 2026

LLM API Pricing: Complete Guide

Everything you need to know about LLM API pricing: how token costs work, what drives pricing differences between models, and a complete price comparison for 18+ models from OpenAI, Anthropic, Google, DeepSeek, and more.

How LLM API Pricing Works

LLM APIs charge by the token — roughly 4 characters of English text, or about 0.75 words. Most providers quote prices per 1 million tokens (per 1M tokens), split into two rates:

Input / Prompt tokens
Tokens in your message to the model — the prompt, context, and any retrieved documents. Input is typically 2–5× cheaper than output.
Output / Completion tokens
Tokens the model generates in its response. More expensive because they require sequential computation that cannot be parallelised.
Quick example: Cost of a single GPT-4.1 request
Prompt: 500 tokens × $2.00/1M = $0.001
Response: 300 tokens × $8.00/1M = $0.0024
Total per request: $0.0034 — about 0.3 cents.

What Drives LLM API Price Differences?

Model size (parameters)

Larger models cost more to run. GPT-4.1 (~200B params) costs $2/1M input; GPT-4.1 Nano (~10B) costs $0.10/1M. The capability–cost tradeoff is steep.

Inference architecture (MoE vs dense)

Mixture-of-Experts models like DeepSeek V4 and GPT-4o activate only a fraction of parameters per token — dramatically reducing compute cost vs dense models at similar quality.

Provider infrastructure & margins

DeepSeek and Mistral pass through lower costs from efficient training and open-weight models. OpenAI and Anthropic charge a premium for proprietary research, safety work, and support.

Context window length

Longer contexts require more KV-cache memory. Models with 1M+ context windows (Gemini, GPT-4.1) may charge more per token on long prompts or use tiered pricing.

Reasoning / thinking tokens

Models like o3, o4-mini, and DeepSeek R1 generate internal "thinking" tokens before answering. These are billed at the same rate — a complex reasoning task may consume 10–100k tokens before you see a word of output.

All LLM APIs — Price Per 1M Tokens (Cheapest First)

ModelProviderInput /1MOutput /1M
Llama 4 ScoutMeta$0.100/1M tkns$0.300/1M tkns
DeepSeek V4 FlashDeepSeek$0.140/1M tkns$0.280/1M tkns
Llama 4 MaverickMeta$0.200/1M tkns$0.800/1M tkns
Gemini 3.1 Flash-LiteGoogle$0.300/1M tkns$2.50/1M tkns
Qwen3.7 PlusAlibaba$0.320/1M tkns$1.28/1M tkns
DeepSeek V4 ProDeepSeek$0.435/1M tkns$0.870/1M tkns
Mistral LargeMistral$0.500/1M tkns$1.50/1M tkns
GPT-5.6 LunaOpenAI$1.00/1M tkns$6.00/1M tkns
Gemini 2.5 ProGoogle$1.25/1M tkns$10.00/1M tkns
Qwen3.7 MaxAlibaba$1.47/1M tkns$4.42/1M tkns
Gemini 3.5 FlashGoogle$1.50/1M tkns$7.50/1M tkns
Mistral Medium 3.5Mistral$1.50/1M tkns$7.50/1M tkns
Claude Sonnet 5Anthropic$2.00/1M tkns$10.00/1M tkns
Grok 4.5xAI$2.00/1M tkns$6.00/1M tkns
GPT-5.6 TerraOpenAI$2.50/1M tkns$15.00/1M tkns
GPT-5.6 SolOpenAI$5.00/1M tkns$30.00/1M tkns
Claude Opus 4.8Anthropic$5.00/1M tkns$25.00/1M tkns
Claude Fable 5Anthropic$10.00/1M tkns$50.00/1M tkns

Prices updated automatically daily. View full comparison table →

LLM API Pricing by Provider

OpenAI

Range: $0.10 (GPT-4.1 Nano) → $15 (o3) per 1M input tokens. GPT-4.1 Mini at $0.40/1M is the sweet spot for most production apps. OpenAI charges a 4× premium on output vs input for most models.

Anthropic

Range: $0.25 (Haiku 3.5) → $15 (Opus 4.7) per 1M input tokens. Claude Sonnet 4.6 at $3/1M is Anthropic's price-performance leader. Anthropic charges 5× output vs input on Claude 3.5+ models.

Google

Range: $0.07 (Gemini 2.0 Flash Lite) → $2.50 (Gemini 2.5 Pro) per 1M input tokens. Gemini 2.5 Flash at $0.30/1M with 1M context is Google's standout value offering.

DeepSeek

Range: $0.07 (V4 Flash) → $0.27 (V4 Pro) per 1M input tokens. DeepSeek provides frontier-quality models at 10–50× lower cost than OpenAI/Anthropic by leveraging MoE architecture and open-weight training.

Frequently Asked Questions

What is the cheapest LLM API in 2026?

DeepSeek V4 Flash at $0.07/1M input tokens is currently the cheapest capable LLM API. For simpler tasks, GPT-4.1 Nano at $0.10/1M and Gemini 2.0 Flash Lite at $0.07/1M are comparable alternatives from major providers.

How much does it cost to run a chatbot using GPT-4?

A chatbot handling 1,000 conversations/day at ~500 tokens each would cost roughly $1/day on GPT-4.1 Mini ($0.40/1M). Using GPT-4.1 at $2/1M input, the same chatbot costs ~$5/day. At 100k conversations/day, costs scale linearly.

Why is Claude more expensive than GPT?

Claude Sonnet 4.6 ($3/1M) is pricier than GPT-4.1 ($2/1M) partly due to Anthropic's Constitutional AI safety research and training costs. However, many developers find Claude's instruction-following justifies the premium for complex tasks.

Do LLM providers charge for cached or repeated tokens?

Many providers offer prompt caching — if you repeat the same system prompt across requests, only the first request pays full price. OpenAI charges 50% for cached inputs; Anthropic charges 10% for cache reads. This can cut costs by 30–70% for apps with large, stable system prompts.

How often do LLM API prices change?

Prices change frequently — often dropping 50–80% per year as providers scale. ComputeCheck updates prices daily via the OpenRouter API. Always verify current pricing on the provider's official page before committing to production architecture.

Ready to compare specific models side-by-side?

Compare AI Model API Pricing →