llmpricingtable.com

What It Is

LLM Pricing Table is a model pricing comparison site for developers and AI application teams, built around the idea of “Compare Up To Date AI Model Costs.” It aggregates providers such as OpenAI, Anthropic, Google Vertex AI, Cohere, DeepSeek, Groq, Mistral, and X.AI, listing input pricing, output pricing, context length, capability tags, and status by model. The page also notes that the data is available as a web table, JSON, CSV, and XML, making it useful for internal spreadsheets or scripted analysis.

Core Capabilities

Its value is not in calling models, but in structuring scattered LLM billing information. The table covers capability tags such as text, audio, image, video, embeddings, TTS, transcription, reasoning, JSON, tool calling, and fine-tuning, while also marking statuses like active, deprecated, and legacy. Typical use cases include model selection, cost estimation, context window comparison, and early-stage vendor screening before procurement. For teams that frequently switch between GPT, Gemini, Claude, Mistral, Groq, or DeepSeek, this kind of aggregation can significantly reduce the time spent checking prices.

Pricing and Integration

The scraped text does not show whether LLM Pricing Table itself has any pricing or subscription model; the site appears to provide direct access to the pricing table. It displays third-party model pricing, with units including /KTok, /MTok, /second, /image, and /character, and also includes information such as cached read pricing and batch processing discounts. On the integration side, JSON, CSV, and XML are highlights, as they can be connected to cost dashboards or automated evaluation workflows. However, there is no clear mention of an official API, SDK, authentication, SLA, or update verification mechanism.

Pros and Cons

Its strengths are broad coverage, straightforward fields, a low barrier to use, and an open GitHub repository for reporting pricing discrepancies and suggesting new providers. The limitations are also clear: it only provides pricing and basic specifications, and does not evaluate model quality, latency, rate limits, regional availability, or stability. The scraped text also shows inconsistencies such as differing units for similar models, N/A, $0.00, or “-”, so final budgeting should still be verified against the vendors’ official pages. There is no clear information about Chinese-language support; the interface content appears to be mainly in English.

Who It’s For and Access from China

It is suitable for AI application developers, technical leads, procurement teams, and FinOps staff who need to quickly build a model cost comparison view. Access from China cannot be determined from the content, so it should be considered unknown. Payments are not the core issue for this site, since actual purchases take place with the individual model providers. Domestic alternatives or complements may include official pricing pages from each provider, the OpenRouter pricing page, Artificial Analysis, the LiteLLM model cost map, or direct comparisons with pricing tables from Chinese large model platforms.