AI Cost CalculatorData Sources & Pricing
Every figure in these tables is taken straight from the provider's official pricing page and re-checked against it. Each source is linked below, and the date we last verified the whole set is shown here: 2026-06-29.
These are the published list prices. Taxes, committed-use or enterprise discounts, and regional rates can still change what lands on your own invoice.
Where the numbers come from
We prefer public pages controlled by the provider: official pricing pages and product documentation. A provider's own billing console or contract takes priority for your account. Blog posts, social announcements, newsletters, and third-party comparison sites may help identify a change, but they are not treated as the final source when an official page is available.
How we choose what to include
- Included models: commonly used text, embedding, image, and open-model serving options that are useful in the calculators.
- Excluded, simplified or missing: many fine-tuning charges, audio and video modalities, enterprise-only terms, promotional credits, taxes, and regional uplifts - and when a separate cached or batch price is not published, the calculator falls back to the standard rate rather than inventing a discount.
A provider may offer several versions of a model with similar names. We keep identifiers separate where the pricing or behavior differs. Retired models may be removed when they are no longer useful for a current comparison.
How rates are normalized
Text and embedding rates are normalized to a price per million tokens where the provider publishes token pricing. Image rates are stored per generated image for the listed configuration. GPU options use an hourly figure and user-supplied throughput rather than pretending hardware has a universal token price.
Input, cached input, and output stay in separate columns. A blank or unavailable discount should not be read as free usage. Batch and cache rules can include minimum lengths, eligible models, retention windows, or processing-time tradeoffs that a single number cannot show.
What the table does not capture
List price is only part of a provider decision. Context-window surcharges, priority processing, data residency, support plans, rate limits, committed-use agreements, and taxes may change the invoice. Tokenization can also differ, so identical text does not always produce identical billable token counts across providers.
Quality and reliability are deliberately not converted into a score here. Test models on your own work and compare cost per accepted result.
Review and correction process
When a rate changes, we update the source data and rebuild the generated tables and calculators. The review date indicates the database review, not a guarantee that every provider left every price unchanged until that moment.
If you spot a discrepancy, send the provider, model, rate, billing condition, and official source URL. Details such as "batch only" or "long context" matter. We review clear corrections and update the database where the source supports them.
Official provider pricing pages
Full text-model pricing table
| Provider | Model | Input /1M | Cached /1M | Output /1M | Context |
|---|---|---|---|---|---|
| OpenAI | GPT-5.5 | $5 | $0.5 | $30 | 400,000 |
| OpenAI | GPT-5.4 | $2.5 | $0.25 | $15 | 400,000 |
| OpenAI | GPT-5.4 mini | $0.75 | $0.075 | $4.5 | 400,000 |
| OpenAI | GPT-5.4 nano | $0.2 | $0.02 | $1.25 | 400,000 |
| OpenAI | GPT-4o | $2.5 | $1.25 | $10 | 128,000 |
| OpenAI | GPT-4o mini | $0.15 | $0.075 | $0.6 | 128,000 |
| Anthropic | Claude Fable 5 | $10 | $1 | $50 | 1,000,000 |
| Anthropic | Claude Opus 4.8 | $5 | $0.5 | $25 | 1,000,000 |
| Anthropic | Claude Sonnet 4.6 | $3 | $0.3 | $15 | 1,000,000 |
| Anthropic | Claude Haiku 4.5 | $1 | $0.1 | $5 | 200,000 |
| Gemini 3.1 Pro | $2 | $0.2 | $12 | 1,000,000 | |
| Gemini 3.5 Flash | $1.5 | $0.15 | $9 | 1,000,000 | |
| Gemini 2.5 Pro | $1.25 | $0.125 | $10 | 1,000,000 | |
| Gemini 2.5 Flash | $0.3 | $0.03 | $2.5 | 1,000,000 | |
| Gemini 2.5 Flash-Lite | $0.1 | $0.01 | $0.4 | 1,000,000 | |
| xAI | Grok 4.3 | $1.25 | N/A | $2.5 | 1,000,000 |
| Mistral | Mistral Medium 3.5 | $1.5 | N/A | $7.5 | 128,000 |
| Mistral | Mistral Large 3 | $0.5 | N/A | $1.5 | 128,000 |
| Mistral | Mistral Small 4 | $0.15 | N/A | $0.6 | 128,000 |
| DeepSeek | DeepSeek-V4-Flash | $0.14 | $0.014 | $0.28 | 1,000,000 |
| DeepSeek | DeepSeek-V4-Pro | $0.44 | $0.044 | $0.87 | 1,000,000 |
| Cohere | Command R+ | $2.5 | N/A | $10 | 128,000 |
| Cohere | Command R | $0.5 | N/A | $1.5 | 128,000 |
| Groq | Llama 3.3 70B (Groq) | $0.59 | N/A | $0.79 | 128,000 |
| Groq | Llama 3.1 8B (Groq) | $0.05 | N/A | $0.08 | 128,000 |
| Together | Llama 3.3 70B (Together) | $0.88 | N/A | $0.88 | 128,000 |
Embedding model pricing
| Provider | Model | Price /1M | Dimensions |
|---|---|---|---|
| OpenAI | text-embedding-3-small | $0.02 | 1,536 |
| OpenAI | text-embedding-3-large | $0.13 | 3,072 |
| Gemini Embedding | $0.15 | 768 | |
| Voyage | voyage-3 | $0.06 | 1,024 |
| Voyage | voyage-3-lite | $0.02 | 512 |
| Cohere | Cohere embed v3 | $0.1 | 1,024 |
| Mistral | mistral-embed | $0.1 | 1,024 |
Image model pricing
| Provider | Model | Price / image |
|---|---|---|
| OpenAI | gpt-image-1 (medium) | $0.04 |
| OpenAI | DALL·E 3 (standard) | $0.04 |
| OpenAI | DALL·E 3 (HD) | $0.08 |
| Black Forest Labs | FLUX.1 Pro | $0.05 |
| Black Forest Labs | FLUX.1 schnell | $0.003 |
| Imagen 3 | $0.04 |
Using these numbers responsibly
Use the tables to shortlist options, then put representative usage into the relevant calculator to see the monthly number. The rates here are the providers' current published prices as of the verified date above, so you can plan against them directly. The only things they can't see are the parts that are specific to your account: tax, enterprise or committed-use discounts, and regional rates.
Spotted an error? Email [email protected] with the model, the disputed figure, its billing condition, and an official source link.