How it works
OpenAI's ChatGPT Plus subscription is neither necessary nor sufficient for using the GPT for Work extensions.
Our self-service pricing option allows you to purchase packs for the amount you need, when you need them. Each pack provides access to all models and features available in GPT for Work.
Available packs
We sell packs of various sizes. Start as low as $29 and buy a bigger pack when you need it.
| Pack price | Tokens included* |
|---|---|
| $29 | Up to 97 million |
| $79 | Up to 260 million |
| $139 | Up to 460 million |
| $299 | Up to 1 billion |
| $599 | Up to 2 billion |
| $999 | Up to 3.3 billion |
* The number of tokens included varies depending on the model used and whether you use your own API key. For example, with the $29 pack, you can get 1.5 million tokens for gpt-4o without an API key, and up to 97 million tokens for gpt-4o-mini with your own API key. Both input and output tokens count towards your token usage. Learn more about our pricing per model. You can also use the cost estimator to estimate what your use case will cost.
What is included in a pack?
Packs include immediate access to:
- All models that do not require an API key.
- All features within the GPT for Sheets, GPT for Excel, and GPT for Word sidebars.
- All models that require an API key if you have set a valid API key. However, packs do NOT include the model provider's API cost for these models.
How packs work
To use GPT for Work extensions, you must have a balance greater than 0*. After consuming your free trial balance, you'll need to buy a GPT for Work pack to increase your balance and continue using the service.
What you need to know about GPT for Work packs:
Balance consumption: Your balance gets consumed more or less quickly depending on the model you use and whether you're using your own API key (see our model price list for details):
Without your own API key: Your balance covers both GPT for Work service cost and the model provider's API cost.
With your own API key: Your balance only covers GPT for Work service cost. You pay the model provider directly for the API cost.
Both input and output tokens count towards your usage. For more information, see how your balance is consumed or use our cost estimator to estimate what your use case will cost.
Expiration: Packs expire one year after the purchase date, and each pack has its own separate expiration date. You can buy more packs at any time to increase your balance.
Cross-platform usage:
Packs bought for Microsoft Excel can be used on Microsoft Word, and vice versa.
Google Sheets packs are separate and can only be used with Google Sheets.
Balance transfer: To transfer balance between Google Sheets and Microsoft Excel/Word, contact our support team.
* Exception: GPT for Docs is free to use but requires an API key. (Tokens are paid directly to the AI provider)
How can I top up my balance?
You can top up your balance and avoid any interruption in service by buying another pack at any time. This increases your balance immediately but doesn't extend the expiration date of your existing packs. Packs are consumed in order of their expiration date, with the pack closest to expiring being used first.
Can I share a pack with my team?
You can add members to your space from the dashboard. Members of a space all share the same balance, which you get from buying packs for the space.
If you need help, contact support.
Model price list
Find the latest model price list for all paid extensions below.
Without an API key
The table below shows the pricing for each model family without an API key (including the model provider's API cost).
| Provider | Model family | Cost | Per |
|---|---|---|---|
| OpenAI | gpt-3.5-turbo | $2.50 | 1M tokens |
| OpenAI | gpt-4o | $20.00 | 1M tokens |
| OpenAI | gpt-4o-mini | $0.90 | 1M tokens |
| OpenAI | gpt-4-turbo | $35.00 | 1M tokens |
| OpenAI | gpt-4 | $65.00 | 1M tokens |
| OpenAI | text-embedding-ada-002 (GPT_MATCH) | $0.25 | 1M tokens |
| gemini-1.5-flash | $2.50 | 1M tokens | |
| Anthropic | claude-3.5-sonnet | $20.00 | 1M tokens |
| Perplexity | llama-3.1-sonar-small-128k-online (GPT_WEB, web browsing bulk tool) | $15.00 | 1K executions |
With an API key
The table below shows the pricing for each family of models or Assistants available with an API key (excluding OpenAI API cost which you pay directly to OpenAI).
| Provider | Model or Assistant | Cost | Per |
|---|---|---|---|
| OpenAI | gpt-3.5-turbo | $1 | 1M tokens |
| OpenAI | gpt-4o | $5 | 1M tokens |
| OpenAI | gpt-4o-mini | $0.3 | 1M tokens |
| OpenAI | gpt-4-turbo | $5 | 1M tokens |
| OpenAI | gpt-4 | $5 | 1M tokens |
| OpenAI | text-embedding-ada-002 (GPT_MATCH) | $0.1 | 1M tokens |
| OpenAI | Assistants | Varies* | 1M tokens |
* Assistants pricing is based on the underlying model of the Assistant used in GPT for Sheets or GPT for Excel. For example, Assistants using gpt-4o cost $5 per million tokens, while those using gpt-4o-mini cost $0.3 per million tokens.
Anthropic models claude-3-haiku, claude-3-opus, claude-3-5-sonnet are free to use with an API key, with API costs paid directly to Anthropic.
Google models gemini-1.5-flash and gemini-1.5-pro are available for GPT for Sheets only, and are currently free to use, excluding API costs which are paid directly to Google.
Mistral AI models mistral-large, mistral-small, ministral-8b, and ministral-3b are available for GPT for Sheets only, and are currently free to use, excluding API costs which are paid directly to Mistral AI.
Models run locally with Ollama are supported for GPT for Excel and GPT for Word, and are currently free to use.
Price evolution
Find below the changes in pricing for all GPT for Work extensions and all models, along with the specific dates on which these changes occurred.
Septembre 10th, 2024 for Excel and Word
Ollama models are now supported by GPT for Excel and Word.
August 29th, 2024 for Excel and Word
Anthropic Claude 3 models are now available for GPT for Excel and Word:
- With an API key: Free to use (API costs paid directly to Anthropic)
- Without an API key: claude-3.5-sonnet available at $20 per 1M tokens
August 7th, 2024 for Excel
GPT for Excel now supports OpenAI Assistants. The pricing is based on the underlying model of the Assistant. For example, Assistants using gpt-4o cost $5 per million tokens, while those using gpt-4o-mini cost $0.3 per million tokens.
August 2nd, 2024 for Sheets
Anthropic's claude-3.5-sonnet model is now available without an API key at $20 per 1M tokens, with an input size limit of 200K tokens.
July 31st, 2024 for Sheets
Assistants pricing is now based on the underlying model of the Assistant, instead of a fixed $5 per million tokens. For example:
- Assistants using gpt-4o: $5 per million tokens
- Assistants using gpt-4o-mini: $0.3 per million tokens
July 24th, 2024 for Excel, and Sheets
- Without an API key: text-embedding-ada-002 (GPT_MATCH) costs $0.25 per 1M tokens, instead of $2.5 previously.
- With an API key: text-embedding-ada-002 (GPT_MATCH) costs $0.1 per 1M tokens, instead of $1 previously.
June 27th, 2024 for Sheets
GPT for Sheets now supports OpenAI Assistants, at a cost of $5 per 1M tokens, with an API key.
May 15th, 2024 for Excel, Word, and Sheets
OpenAI gpt-4o model is now available for $20 per 1M tokens without an API key, and $5 per 1M tokens with an API key.
April 4th, 2024 for Excel, Word, and Sheets
GPT for Excel and Word prices have dropped for all models:
Without an API key (including OpenAI and Azure API cost):
- gpt-3.5 turbo models cost $2.5 per 1M tokens, instead of $25 previously
- gpt-4-turbo models cost $35 per 1M tokens, instead of $55 previously
- gpt-4 models cost $65 per 1M tokens, instead of $85 previously
- text-embedding-ada-002 costs $2.5 for 1M tokens, instead of $25 previously
- Perplexity model costs $15 per 1K executions, instead of $50 previously
With an API key (excluding OpenAI API cost which you pay directly to OpenAI):
- gpt-3.5 turbo models cost $1 per 1M tokens, instead of $15 previously
- gpt-4 and gpt-4-turbo models cost $5 per 1M tokens, instead of $15 previously
- text-embedding-ada-002 costs $1 per 1M tokens, instead of $15 previously
GPT for Sheets prices have been simplified:
The cost of input and output tokens is now identical.
Without an API key (including OpenAI and Azure API cost):
- gpt-3.5 turbo models cost $2.5 per 1M tokens, instead of $6.00 per 1M input tokens and $8.00 per 1M output tokens for Azure models and $2.00 per 1M input tokens and $6.00 per output tokens for OpenAI models
- gpt-4-turbo models cost $35 per 1M tokens, instead of $40.00 per 1M input tokens and $120.00 per 1M output tokens
- gpt-4 models cost $65 per 1M tokens, instead of $120.00 per 1M input tokens and $240.00 per 1M output tokens for Azure models
- text-embedding-ada-002 costs $2.5 per 1M tokens, instead of $0.40 per 1M input tokens
- Perplexity model costs $15 per 1K executions, instead of $50 previously
With an API key (excluding OpenAI API cost which you pay directly to OpenAI):
- gpt-3.5 turbo models cost $1 per 1M tokens, instead of $1.50 per 1M input tokens and $2.00 per 1M output tokens
- gpt-4 models cost $5 per 1M tokens, instead of $30.00 per 1M input tokens and $60.00 per 1M output tokens
- gpt-4-turbo models cost $5 per 1M tokens, instead of $10.00 per 1M input tokens and $30.00 per 1M output tokens
- text-embedding-ada-002 costs $1 per 1M tokens, instead of $0.40 for 1M input tokens
March 13th, 2024 for GPT for Excel and Word
The cost of input and output tokens is now identical.
Without an API key (including OpenAI and Azure API cost):
- gpt-3.5-turbo models now cost $25 per 1M tokens, instead of $6.00 per 1M input tokens and $8.00 per 1M output tokens for Azure models and $2.00 per 1M input tokens and $6.00 per output tokens for OpenAI models
- gpt-4 models now cost $85 per 1M tokens, instead of $120.00 per 1M input tokens and $240.00 per 1M output tokens for Azure models
- gpt-4-turbo models now cost $55 per 1M tokens, instead of $40.00 per 1M input tokens and $120.00 per 1M output tokens
- text-embedding-ada-002 now costs $25 per 1M tokens, instead of $0.40 per 1M input tokens
With an API key, the pricing is the same across all models at $15 per 1M tokens (OpenAI API cost is not included as you pay it directly to OpenAI through your API key).
These prices only apply to packs bought after March 13th, 2024. Packs bought on GPT for Excel before March 13th, 2024 continue to be consumed at the prices that were listed when they were bought.
Until March 13th, 2024 for GPT for Excel and Sheets
Without an API key (including OpenAI and Azure API cost):
- gpt-3.5-turbo Azure models cost $6.00 per 1M input tokens and $8.00 per 1M output tokens
- gpt-4 Azure models cost $120.00 per 1M input tokens and $240.00 per 1M output tokens
- gpt-4-turbo models cost $40.00 per 1M input tokens and $120.00 per 1M output tokens
- gpt-3.5-turbo OpenAI models cost $2.00 per 1M input tokens and $6.00 per output tokens
- text-embedding-ada-002 (with GPT_MATCH) costs $0.40 for 1M input tokens
With an API key (excluding OpenAI API cost which you pay directly to OpenAI):
- gpt-3.5-turbo models cost $1.50 per 1M input tokens and $2.00 per 1M output tokens
- gpt-4-turbo models cost $10.00 per 1M input tokens and $30.00 per 1M output tokens
- gpt-4 models cost $30.00 per 1M input tokens and $60.00 per 1M output tokens
- gpt-4-1106-vision-preview (with GPT_VISION) costs $10.00 per 1M input tokens and $30.00 per 1M output tokens
- text-embedding-ada-002 (with GPT_MATCH) costs $0.10 for 1M input tokens.
- gpt-3.5-turbo-instruct models cost $1.50 per 1M input tokens and $2.00 per 1M output tokens
- gpt-3.5-turbo (fine-tuned) models cost $3.00 per 1M input tokens and $6.00 per 1M output tokens
- babbage-002 (fine-tuned) models cost $1.60 per 1M input tokens and $1.60 per 1M output tokens
- davinci-002 (fine-tuned) models cost $12.00 per 1M input tokens and $12.00 per 1M output tokens