Question 1

What is LLMPrice?

Accepted Answer

LLMPrice is a free LLM API pricing comparison tool built on LiteLLM's model_prices_and_context_window.json, covering OpenAI, Anthropic, Google, DeepSeek, and many other providers.

Question 2

Where can I check GPT, Claude, and Gemini API prices?

Accepted Answer

The homepage lists popular text models. Open /text-all to search or filter the full catalog by provider and view input price, output price, and context window, normalized to cost per million tokens.

Question 3

How often is pricing data updated?

Accepted Answer

Pricing is fetched from LiteLLM's upstream JSON with about 1 hour caching on the server and in your browser. Updates propagate on the next refresh cycle.

Question 4

Which currencies are supported?

Accepted Answer

USD, CNY, EUR, JPY, HKD, GBP, KRW, TWD, SGD, and AUD. Prices originate in USD and convert using live exchange rates.

Question 5

What is the difference between input and output pricing?

Accepted Answer

Input price covers prompt tokens sent to the model. Output price covers tokens generated in the response. Output tokens are often priced higher than input tokens.

Question 6

Are these prices exactly the same as vendor websites?

Accepted Answer

Data comes from LiteLLM's community-maintained config and may lag official vendor pages. Tiered pricing shows the first tier only — always confirm on the provider's official docs.

Question 7

How do I compare DeepSeek, Qwen, or Kimi pricing?

Accepted Answer

Use the search bar or provider filter for dashscope (Qwen), deepseek, moonshot (Kimi), etc., then sort by input or output price to compare side by side.

Question 8

How are AI image generation models priced — per image or per token?

Accepted Answer

It varies by vendor: OpenAI GPT Image 2 charges per token (image output $30/M tokens); FLUX.1 Pro charges per megapixel (MP); Chinese models like Tongyi Wanxiang, Doubao, and CogView mostly charge per image (¥0.06–¥0.5/image). The /image page lists API pricing and subscription plans side by side, filterable by channel.

Question 9

What's the per-second cost of AI video generation models?

Accepted Answer

720p per-second costs vary widely: Tongyi wan2.2-i2v-flash ~¥0.1/s (480P), Veo 3.1 Lite $0.05/s, Kling 2.6 no-audio $0.07/s, Veo 3.1 Standard $0.40/s, Sora 2 Pro 1024p $0.50/s. The /video page lists both API and subscription channels, sortable by price.

Question 10

Should I use the API or a subscription for image/video models?

Accepted Answer

API pay-as-you-go suits developers integrating generation into products; subscriptions suit creators who want a fixed monthly fee with included quota. The image/video pages offer All / API·Official / API·Relay / Subscription filters to compare both acquisition channels for the same model.

Popular text model prices