Anthropic's developer API for the Claude model family, billed per million tokens with separate input, output and prompt-caching rates.
The Claude API gives developers programmatic access to Claude models for chat, analysis, coding and agent applications. Pricing is per million tokens (MTok), with different input and output rates per model.
Prompt caching has its own write and read rates; cache hits are priced as a fraction of base input. The pricing page notes some models are limited availability or retired, so check the model list before choosing.
Usage-based pricing
| Item | Rate | Details | Source and check date |
|---|---|---|---|
| Claude Fable 5.1, input | USD 10 per 1M input tokens | Cache hits priced at $0.25 per MTok (0.025x base); 5m cache write $12.50, 1h cache write $20 per MTok. | Source for Claude Fable 5.1, input (opens in a new tab)Checked October 24, 2026 |
| Claude Fable 5.1, output | USD 50 per 1M output tokens | Source for Claude Fable 5.1, output (opens in a new tab)Checked October 24, 2026 | |
| Claude Opus 5.5, input | USD 4 per 1M input tokens | Cache hits priced at $0.20 per MTok (0.05x base); 5m cache write $5, 1h cache write $8 per MTok. | Source for Claude Opus 5.5, input (opens in a new tab)Checked October 24, 2026 |
| Claude Opus 5.5, output | USD 20 per 1M output tokens | Source for Claude Opus 5.5, output (opens in a new tab)Checked October 24, 2026 | |
| Claude Sonnet 5.5, input | USD 2 per 1M input tokens | Cache hits priced at $0.10 per MTok (0.05x base); 5m cache write $2.50, 1h cache write $4 per MTok. | Source for Claude Sonnet 5.5, input (opens in a new tab)Checked October 24, 2026 |
| Claude Sonnet 5.5, output | USD 10 per 1M output tokens | Source for Claude Sonnet 5.5, output (opens in a new tab)Checked October 24, 2026 |
Prices checked on October 24, 2026 from the vendor's official page. Vendors change prices; confirm before buying. How we check prices. Official pricing page (opens in a new tab)
Developers and companies building applications and agents on Claude models.
This page is based on desk research of the vendor's public pages and documentation. SIDAN Stack has not tested this product hands-on. How we research. Last verified: October 24, 2026.
Hugging Face Hub is a platform for hosting and sharing machine learning models, datasets and demo apps (Spaces) in Git-based repositories. It offers model and dataset cards, a dataset viewer, Inference Providers for serverless model calls, Inference Endpoints for dedicated deployments, Spaces hardware including ZeroGPU, and organization features such as SSO, audit logs and storage regions on Team and Enterprise plans.
Free plan available; paid plans from USD 9 per month
Mistral AI's developer API (Studio) for its language, OCR, speech, embedding and moderation models, billed per token, per page or per minute depending on the service, with batch and cached-input discounts.
Usage-based pricing
OpenAI's developer API for calling its language models, billed per million tokens with separate input, cached-input and output rates.
Usage-based pricing
Pinecone is a managed vector database for AI applications. Serverless indexes support dense, sparse, full-text and hybrid search with metadata filtering, namespaces for multi-tenancy, integrated embedding and reranking models, backups and dedicated read nodes. Pinecone Assistant adds a managed RAG service. A free Starter plan is available; paid plans bill on storage, read and write units.
Free plan available; paid plans from USD 20 per month
To compare AI API pricing, estimate input, output and cached-input tokens per request, copy each model's per-million-token rates with their tier and context band, multiply by monthly volume, apply only documented batch and caching discounts, and check the result against the provider's rate limits and spend caps. This tutorial walks through the method with verified OpenAI and Anthropic rates as listed on 24 October 2026 and an illustrative worked example.