Anthropic API vs OpenAI API — Pricing

Side-by-side pricing comparison of Anthropic's Claude API and OpenAI's API in 2026. Per-tier costs, cache discounts, batch savings, and which to pick.

🔥 Launch tonight — Power Prompts PDF 50p (just 50p tonight)30 battle-tested Claude Code prompts · 8 pages · paste into CLAUDE.md · price reverts to £5

Anthropic (Claude) and OpenAI (GPT) are the two API providers most teams evaluate side-by-side in 2026. Both have three tiers, both bill per token, both offer caching and batch discounts. Their pricing structures look similar at first glance — but the per-tier cost ratios differ meaningfully.

Anthropic tiers (per million tokens, 2026)

TierModelInputOutput
TopClaude Opus 4.7$15$75
MidClaude Sonnet 4.6$3$15
LowClaude Haiku 4.5$1$5

OpenAI tiers (publicly listed reference rates, 2026)

OpenAI's lineup separates "reasoning" models from "chat" models. Pricing varies by family; consult the OpenAI pricing page for live numbers. As a rough comparison, OpenAI's top-tier reasoning models are priced in the same band as Opus 4.7 (~$15–$60 per million input depending on the model), while their mid-tier chat models compete with Sonnet 4.6 at roughly $3–$10 per million input.

Where the structures differ

Which to pick?

For Claude-side estimates, use the Claude Cost Calculator. For multi-model comparison including OpenAI, see the LLM Prompt Pricing tool.

Frequently asked questions

Is Anthropic's Claude API cheaper than OpenAI's API?
It depends on tier and workload. For workloads with heavy prompt-cache reuse (RAG, repeated long-context), Anthropic's 90%-off cached-read pricing makes it markedly cheaper than OpenAI. For single-shot generation without caching, the two are within ~20% of each other at comparable tiers.
Do Anthropic and OpenAI offer batch pricing?
Yes — both offer roughly 50% off standard pricing for batched async requests with a 24h SLA. Use batch for offline classification, eval runs, or any workload where you don't need real-time response.
Can I mix Claude and OpenAI in the same product?
Yes. Many production apps route by task: Claude for long-context summarization and coding, OpenAI for specific reasoning tasks. The Vercel AI SDK, LangChain, and LiteLLM all support both providers behind a unified interface.

Free tools

Cost Calculator → Prompt-Pricing Recommender → Diff Summarizer → Skills Browser →

Related

Claude Opus 4.7 vs Sonnet 4.6 Pricing (2026 Comparison)How Much Does Claude Cost? (2026 API Pricing Guide)Claude Prompt Caching: 90% Cost Savings Explained (2026)Claude API Cost Calculator: Estimate Your Anthropic BillClaude vs GPT-4 Pricing: 2026 API Cost Comparison