Everything you need to launch and grow your SaaS
Calculate and compare the cost of using OpenAI, Azure, Anthropic Claude, Llama 2, Google Gemini, Mistral, and Cohere LLM APIs for your AI project with our simple and powerful free calculator. Latest numbers as of August 25, 2024
| Provider | Model | Knowledge Cutoff | Context | Input/1M tokens | Output/1M tokens | Per Call | Chat/Completion Models |
|---|---|---|---|---|---|---|---|
| Mistral (via Anyscale) | mistral-7b-instruct-v0.1 | Sep 2021 | 32K | $0.15 | $0.15 | $0.0001 | $0.01 |
| OpenAI / Azure | gpt-4o-mini | October 2023 | 128K | $0.15 | $0.60 | $0.0001 | $0.01 |
| Mistral AI | open-mistral-7b | 32K | $0.25 | $0.25 | $0.0002 | $0.02 | |
| Amazon | titan text lite | N/A | 4K | $0.30 | $0.40 | $0.0002 | $0.02 |
| gemini 1.5 flash | Early 2023 | 1 Million | $0.35 | $0.70 | $0.0002 | $0.02 | |
| Anthropic | claude haiku | Aug 2023 | 200K | $0.25 | $1.25 | $0.0003 | $0.03 |
| OpenAI / Azure | gpt-3.5-turbo-0125 | Sep 2021 | 16K | $0.50 | $1.50 | $0.0004 | $0.04 |
| gemini 1.0 pro | Early 2023 | 32K | $0.50 | $1.50 | $0.0004 | $0.04 | |
| Mistral AI | open-mixtral-8x7b | Sep 2021 | 32K | $0.70 | $0.70 | $0.0004 | $0.04 |
| Amazon | titan text express | N/A | 8K | $0.80 | $1.60 | $0.0006 | $0.06 |
| Meta (via Anyscale) | llama 2 70b | Sep 2022 | 4K | $1.00 | $1.00 | $0.0006 | $0.06 |
| OpenAI / Azure | gpt-3.5-turbo-instruct | Sep 2021 | 4K | $1.50 | $2.00 | $0.0010 | $0.10 |
| Mistral AI | mistral-small-latest | Sep 2021 | 32K | $2.00 | $6.00 | $0.0016 | $0.16 |
| Mistral AI | mistral-medium-latest | Sep 2021 | 32K | $2.70 | $8.10 | $0.0022 | $0.22 |
| gemini 1.5 pro | Early 2023 | 1 Million | $3.50 | $10.50 | $0.0028 | $0.28 | |
| Anthropic | claude sonnet | Aug 2023 | 200K | $3.00 | $15.00 | $0.0030 | $0.30 |
| OpenAI / Azure | gpt-4o | October 2023 | 128K | $5.00 | $15.00 | $0.0040 | $0.40 |
| OpenAI / Azure | gpt-4o-2024-05-13 | October 2023 | 128K | $5.00 | $15.00 | $0.0040 | $0.40 |
| Mistral AI | mistral-large-latest | Sep 2021 | 32K | $8.00 | $24.00 | $0.0064 | $0.64 |
| OpenAI / Azure | gpt-4-0125-preview | Dec 2023 | 128K | $10.00 | $30.00 | $0.0080 | $0.80 |
| OpenAI / Azure | gpt-4-1106-preview | Apr 2023 | 128K | $10.00 | $30.00 | $0.0080 | $0.80 |
| OpenAI / Azure | gpt-4-1106-vision-preview | Apr 2023 | 128K | $10.00 | $30.00 | $0.0080 | $0.80 |
| Anthropic | claude opus | Aug 2023 | 200K | $15.00 | $75.00 | $0.0150 | $1.50 |
| OpenAI / Azure | gpt-4 | Sep 2021 | 8K | $30.00 | $60.00 | $0.0210 | $2.10 |
| OpenAI / Azure | gpt-4-32k | Sep 2021 | 32k | $60.00 | $120.00 | $0.0420 | $4.20 | Fine-tuning Models |
| palm 2 | Mid 2021 | 8K | $2.00 | $2.00 | $0.0012 | $0.12 | |
| OpenAI | gpt-3.5 turbo | Sep 2021 | 4K | $3.00 | $6.00 | $0.0021 | $0.21 | Embedding Models |
| OpenAI / Azure | text-embedding-3-small | $0.02 | $0.0000 | $0.00 | |||
| OpenAI / Azure | ada v2 | $0.10 | $0.0001 | $0.01 | |||
| Cohere | embed | $0.10 | $0.0001 | $0.01 | |||
| Mistral | embed | $0.10 | $0.0001 | $0.01 | |||
| Amazon | tital embeddings | $0.10 | $0.0001 | $0.01 | |||
| OpenAI / Azure | text-embedding-3-large | $0.13 | $0.0001 | $0.01 | |||
| palm 2 | $0.40 | $0.0002 | $0.02 |
Hover for secret coupon code
NOWORNEVER
$100 off, no subscripiton!
Distribution first SaaS boilerplate
The SvelteKit(frontend) + NestJs(backend) boilerplate to build your next SaaS, Web App, GPT Actions or AI Tools and become profitable!
You can find the official OpenAI pricing here .
For Mistral Platform pricing and rate limits, you can find them here .
Claude AI pricing and rate limits can be found here .
Google’s AI pricing, or gemini pricing, or palm pricing can be found here .
Amazon’s AI pricing, or bedrock pricing, can be found here .
A token is a part of a text that can be word, subword, punctuation mark or symbol. It is the unit of account of OpenAI APIs.
An OpenAI execution is the combination of the invite to OpenAI and the response from OpenAI. All text in the prompt and response count for OpenAI billing.
The OpenAI prompt is the instructions or question you send to OpenAI in order to get a response. It is text written in normal, natural language, for example: Prompt: “Write a tagline for an ice cream shop”.
This is the response OpenAI gives you to the prompt you sent, for example: Prompt: “Write a tagline for an ice cream shop” Response: “We serve up smiles with every scoop!”
You can configure a billing limit in the OpenAI billing settings , in order to control your costs.
You can review your usage in the OpenAI dashboard usage section. You can easily set a spending limit in the OpenAI dashboard billing section.
You can use OpenAI’s tokenizer to estimate your token usage for a given prompt. You can find the calculator here .
Get actionable tips and insights on how to save on your AI project.