Which LLM API is cheapest in 2026?

DeepSeek and Cerebras offer the lowest per-token costs for most workloads. Use the calculator with your actual token volume to find the cheapest option for your input/output ratio.

What is batch pricing for LLM APIs?

Batch pricing offers up to 50% discount on calls that do not require real-time responses. OpenAI, Anthropic, and Google support batch mode. Toggle the batch discount to see the savings.

What is cached input token pricing?

Cached tokens reduce cost for repeated system prompts. Anthropic and OpenAI charge 50-90% less for cached tokens. Enable the cached pricing toggle to calculate savings.

What export formats are available?

PDF export is free. With a free account you unlock CSV, HTML, Markdown, LiteLLM config, OpenRouter JSON, Python Dict, .env Snippet, Plain Text, and Cursor Rules. 10 formats total.

Does the calculator show context window sizes?

Yes. Every model shows its context window in thousands of tokens (K). Useful when choosing a model for long documents or multi-turn conversations.

What is the LiteLLM export format?

LiteLLM is a Python library for unified LLM API access. The export generates a ready-to-use model_list config you can paste directly into your LiteLLM or Omnirouter setup.

How often is LLM pricing updated?

Pricing is updated whenever providers change their rates. Each entry shows a last-verified date. For production decisions, cross-check against the provider official pricing page.

ComparEdge

LLM API Pricing Calculator

Compare API costs across OpenAI, Anthropic, Google, DeepSeek, and more. Enter your daily token usage and instantly see which model fits your budget.

17 providers70 modelsBatch discountsCached pricing

Pricing data updated 2026-07-10

More free tools

Professional-grade tools for SaaS decision makers.

Tech Stack Builder

Build your SaaS stack from 44 categories and get a full cost and integration analysis.

Vendor Lock-In Checker

Score vendor lock-in risk across 5 dimensions before signing any contract.

Cyber Risk Calculator

Estimate data breach exposure using IBM methodology across 7 security domains.

Everything you need to pick the right LLM API

Built for developers and AI teams who care about cost at scale.

70 Models, 17 Providers

OpenAI, Anthropic, Google, DeepSeek, Groq, Mistral, xAI, Cohere, and more. Always up to date.

Batch Discount Toggle

50% off for eligible providers when real-time response is not required. One click to compare.

Cached Token Pricing

See real costs with prompt caching enabled. Cuts input token cost by 50-90% on repeated context.

Compare Mode

Pin two model sets side by side. See cost delta and context window gap for your exact workload.

10 Export Formats

PDF free. CSV, Python dict, LiteLLM config, OpenRouter JSON, .env snippet, Cursor Rules and more with free account.

Save and Compare History

Free account to save calculations, revisit past results, and track how pricing changes over time.

LLM API Pricing Reference (2026)

Prices per 1 million tokens. Use the calculator above to estimate your exact monthly cost.

Provider	Model	Input / 1M tokens	Output / 1M tokens	Context	Tier
OpenAI	GPT-5.6 Sol	$5	$30	1000K	Flagship
OpenAI	GPT-5.6 Luna	$1	$6	1000K	Fast
Anthropic	Claude Opus 4.8	$5	$25	1000K	Flagship
Anthropic	Claude Haiku 4.5	$1	$5	200K	Fast
Google	Gemini 3.1 Pro	$2	$12	1000K	Flagship
Google	Gemini 3.5 Flash	$1.5	$9	1000K	Fast
xAI	Grok 4.5	$2	$6	500K	Reasoning
DeepSeek	DeepSeek V4 Pro	$0.435	$0.87	1000K	Economy
DeepSeek	DeepSeek V4 Flash	$0.14	$0.28	1000K	Economy
Moonshot	Kimi K2.6	$0.95	$4	256K	Flagship

Prices verified 2026-07-10. For 70 models across 17 providers with custom token volume, use the interactive calculator above.

10 Export Formats

PDF is free. Unlock the rest with a free account.

FREE

PDF Report

HTML File

CSV Spreadsheet

Plain Text

Markdown Table

LiteLLM Config

OpenRouter JSON

Python Dict

.env Snippet

Cursor Rules

Frequently Asked Questions

Click any question to expand the answer.