LLM API pricing usually depends on the selected model, normal input tokens, cached input where supported, and generated output tokens. Providers publish separate USD rates per one million tokens. The table below compares standard direct-API text rates; actual workload cost also depends on output length, retries, tools, media, regional service, and platform terms.
This LLM API pricing table lists standard direct-API USD prices per one million text tokens for 17 current GPT, Claude, Gemini, and DeepSeek models. Long-context alternatives appear beneath affected rates, and the workload calculator applies them automatically. Each row links to its official provider source and shows the verified date. A missing value is unavailable, never assumed to be zero.
| Provider / model | Input / 1M | Cached / 1M | Output / 1M | Context | Accuracy | Verified |
|---|---|---|---|---|---|---|
| Claude Fable 5Anthropic | $10 | $1 | $50 | 1,000,000 | Estimate | 2026-08-09Verified |
| Claude Haiku 4.5Anthropic | $1 | $0.1 | $5 | 200,000 | Estimate | 2026-08-09Verified |
| Claude Opus 5Anthropic | $5 | $0.5 | $25 | 1,000,000 | Estimate | 2026-08-09Verified |
| Claude Sonnet 4.6Anthropic | $3 | $0.3 | $15 | 1,000,000 | Estimate | 2026-08-09Verified |
| Claude Sonnet 5Anthropic | $2 | $0.2 | $10 | 1,000,000 | Estimate | 2026-08-09VerifiedValid through 2026-08-31 |
| DeepSeek V4 FlashDeepSeek | $0.14 | $0.0028 | $0.28 | 1,000,000 | Estimate | 2026-08-09Needs review |
| DeepSeek V4 ProDeepSeek | $0.435 | $0.003625 | $0.87 | 1,000,000 | Estimate | 2026-08-09Needs review |
| Gemini 2.5 FlashGoogle | $0.3 | $0.03 | $2.5 | 1,048,576 | Estimate | 2026-08-09Verified |
| Gemini 3.1 Pro PreviewGoogle | $2Over 200K: $4 | $0.2Over 200K: $0.4 | $12Over 200K: $18 | 1,048,576 | Estimate | 2026-08-09Verified |
| Gemini 3.5 FlashGoogle | $1.5 | $0.15 | $9 | 1,048,576 | Estimate | 2026-08-09Verified |
| Gemini 3.5 Flash-LiteGoogle | $0.3 | $0.03 | $2.5 | 1,048,576 | Estimate | 2026-08-09Verified |
| Gemini 3.6 FlashGoogle | $1.5 | $0.15 | $7.5 | 1,048,576 | Estimate | 2026-08-09Verified |
| GPT-5.4 miniOpenAI | $0.75 | $0.075 | $4.5 | 400,000 | Compatible | 2026-08-09Verified |
| GPT-5.4 nanoOpenAI | $0.2 | $0.02 | $1.25 | 400,000 | Compatible | 2026-08-09Verified |
| GPT-5.6 LunaOpenAI | $0.2Over 272K: $0.4 | $0.02Over 272K: $0.04 | $1.2Over 272K: $1.8 | 1,050,000 | Compatible | 2026-08-09Verified |
| GPT-5.6 SolOpenAI | $5Over 272K: $10 | $0.5Over 272K: $1 | $30Over 272K: $45 | 1,050,000 | Compatible | 2026-08-09Verified |
| GPT-5.6 TerraOpenAI | $2Over 272K: $4 | $0.2Over 272K: $0.4 | $12Over 272K: $18 | 1,050,000 | Compatible | 2026-08-09Verified |
The 17 model rows use USD list prices per one million text tokens, checked against the linked official provider pages. OpenAI and Gemini long-context tiers are applied automatically from the entered input count. Batch, audio, image, video, tools, storage, regional, and platform fees may differ.
Start with the 5 million input and 1 million output token reference workload, then replace those values with your own input, cached input, output, and request volume. Switch the selected model to test another row from the complete comparison above; the result is conditional on those assumptions and the currently verified rates.
This estimate uses published text-token rates. It excludes tool calls, media, storage, taxes, regional pricing, platform fees, and provider-side token differences.
Every LLM API pricing row was checked on against the linked OpenAI, Anthropic, Google, or DeepSeek pricing page. Recheck the LLM API pricing source before a purchasing decision. Claude Sonnet 5 shows the official introductory rate and its August 31, 2026 valid-through date. DeepSeek LLM API pricing remains flagged for review because its official page warns of a planned price increase.
Output often uses a higher per-million rate than input. A model with a low input price can produce a higher total when answers are long, retries are common, or a workflow generates large structured results.
Compare typical and high-output cases before choosing a model from the LLM API pricing table.
LLM API pricing for cached input applies only when a provider recognizes an eligible cache hit. Writes, storage, retention, minimum prefix size, and invalidation can change the economics. The calculator models reads, not automatic eligibility.
No. In LLM API pricing, “Lowest estimated cost” describes one selected workload. Quality, latency, context, tool support, data handling, availability, and engineering fit can outweigh the token line item.
Use the GPT vs Claude API cost comparison when those two providers are the shortlist. Use the dedicated GPT or Claude calculator for model-family detail.
The LLM API pricing comparison checked 17 active model rows against current OpenAI, Anthropic Claude, Google Gemini, and DeepSeek first-party pages, then measured identical token workloads with one cost formula and applicable long-context rate tiers. Its limitation is commercial scope: batch, media, tools, storage, regional service, taxes, credits, and partner-platform pricing are excluded. OpenAI official pricing. Last verified: .
| Evidence | Observed result or boundary |
|---|---|
| Verified model rows | 17 models across OpenAI, Anthropic Claude, Google Gemini, and DeepSeek. |
| Measured workload | Normal input, cached input, output, request volume, monthly and annual cost. |
LLM API pricing questions
This LLM API pricing table was checked against each linked official provider page on 2026-08-09. DeepSeek is marked Needs review because its official page still announces a planned price increase.
No. This LLM API pricing table compares standard paid direct-API text rates. Free quota, promotional credits, batch discounts, enterprise contracts, and platform fees are separate.
The lowest LLM API pricing estimate depends on the selected models, input, cache, output, and request volume. The sorted result applies only to that workload and current published rates.
No verified unified affiliate provider is active in the LLM API pricing table. Any future partner must have a real URL, rights review, disclosure, and independently checked terms.
This LLM API pricing comparison is suitable for screening a defined text workload, but it is not for creating a universal model ranking. Use realistic input, output, cache, and retry data, then recheck the official source before a purchase or deployment decision.