How credits are consumed
Credits are consumed based on your actual model usage, which consists of tokens sent and received, certain tool executions such as web searches, and the number of requests you make. Small and short tasks cost little, while bigger and longer tasks cost more. See detailed prices below.
Agent models​
Agent models handle the Agent's chat messages, thinking, and spreadsheet tasks that do not require bulk processing with AI.
- Subscription
- Pay-as-you-go
Built-in models​
| Provider | Model | Input | Cached input | Cache write | Output | Request | Average cost |
|---|---|---|---|---|---|---|---|
 OpenAI | GPT-5.6 Sol | $7.14 / 1M tokens | $0.72 / 1M tokens | $8.93 / 1M tokens | $42.86 / 1M tokens | - | N/A |
| OpenAI | GPT-5.6 Luna | $2.86 / 1M tokens | $0.29 / 1M tokens | $3.57 / 1M tokens | $17.14 / 1M tokens | - | N/A |
| OpenAI | GPT-5.5 | $7.14 / 1M tokens | $0.72 / 1M tokens | - | $42.86 / 1M tokens | - | $41 / 100 messages |
| OpenAI | GPT-5.4 | $3.57 / 1M tokens | $0.36 / 1M tokens | - | $21.43 / 1M tokens | - | $25 / 100 messages |
| OpenAI | GPT-5.1 | $3.57 / 1M tokens | $0.36 / 1M tokens | - | $28.57 / 1M tokens | - | $8 / 100 messages |
 Anthropic | Opus 4.8 | $7.14 / 1M tokens | $0.72 / 1M tokens | $8.93 / 1M tokens | $35.72 / 1M tokens | - | $65 / 100 messages |
| Anthropic | Opus 4.7 | $7.14 / 1M tokens | $0.72 / 1M tokens | $8.93 / 1M tokens | $35.72 / 1M tokens | - | N/A |
| Anthropic | Opus 4.6 | $7.14 / 1M tokens | $0.72 / 1M tokens | $8.93 / 1M tokens | $35.72 / 1M tokens | - | $43 / 100 messages |
| OpenAI | Web search (gpt-5.4-mini) | $1.07 / 1M tokens | $0.11 / 1M tokens | - | $6.43 / 1M tokens | $14.29 / 1K requests | N/A |
Average costs are based on recent usage data. N/A means that no benchmark has been measured yet.
Provider-connected models​
| Provider | Model | Input | Cached input | Cache write | Output | Request | Average cost |
|---|---|---|---|---|---|---|---|
| Varies | All provider-connected models | - | - | - | - | $1 / 1K requests | N/A |
Average costs are based on recent usage data. N/A means that no benchmark has been measured yet.
Built-in models​
| Provider | Model | Input | Cached input | Cache write | Output | Request | Average cost |
|---|---|---|---|---|---|---|---|
| OpenAI | GPT-5.6 Sol | $12 / 1M tokens | $1.20 / 1M tokens | $15 / 1M tokens | $51 / 1M tokens | - | N/A |
| OpenAI | GPT-5.6 Luna | $8 / 1M tokens | $0.80 / 1M tokens | $9 / 1M tokens | $36 / 1M tokens | - | N/A |
| OpenAI | GPT-5.5 | $12 / 1M tokens | $1.20 / 1M tokens | - | $51 / 1M tokens | - | $66 / 100 messages |
| OpenAI | GPT-5.4 | $9 / 1M tokens | $0.90 / 1M tokens | - | $42 / 1M tokens | - | $55 / 100 messages |
| OpenAI | GPT-5.1 | $7 / 1M tokens | $0.70 / 1M tokens | - | $42 / 1M tokens | - | $19 / 100 messages |
| Anthropic | Opus 4.8 | $10 / 1M tokens | $1 / 1M tokens | $12.50 / 1M tokens | $42 / 1M tokens | - | $145 / 100 messages |
| Anthropic | Opus 4.7 | $10 / 1M tokens | $1 / 1M tokens | $12.50 / 1M tokens | $42 / 1M tokens | - | $99 / 100 messages |
| Anthropic | Opus 4.6 | $10 / 1M tokens | $1 / 1M tokens | $12.50 / 1M tokens | $42 / 1M tokens | - | $198 / 100 messages |
| OpenAI | Web search (gpt-5.4-mini) | $2 / 1M tokens | $0.20 / 1M tokens | - | $12 / 1M tokens | $20 / 1K searches | N/A |
Average costs are based on recent usage data. N/A means that no benchmark has been measured yet.
Provider-connected models​
| Provider | Model | Input | Cached input | Cache write | Output | Request | Average cost |
|---|---|---|---|---|---|---|---|
| OpenAI | GPT-5.6 Sol | $7 / 1M tokens | $0.70 / 1M tokens | $8.75 / 1M tokens | $21 / 1M tokens | - | N/A |
| OpenAI | GPT-5.6 Luna | $7 / 1M tokens | $0.70 / 1M tokens | $6.50 / 1M tokens | $30 / 1M tokens | - | N/A |
| OpenAI | GPT-5.5 | $7 / 1M tokens | $0.70 / 1M tokens | $0 / 1M tokens | $21 / 1M tokens | - | $34 / 100 messages |
| OpenAI | GPT-5.4 | $4 / 1M tokens | $0.40 / 1M tokens | $0 / 1M tokens | $12 / 1M tokens | - | $17 / 100 messages |
| OpenAI | GPT-5.1 | $4.50 / 1M tokens | $0.45 / 1M tokens | $0 / 1M tokens | $22 / 1M tokens | - | $10 / 100 messages |
| Anthropic | Fable 5 | $5 / 1M tokens | $0.50 / 1M tokens | $6.25 / 1M tokens | $17 / 1M tokens | - | N/A |
| Anthropic | Opus 4.8 | $5 / 1M tokens | $0.50 / 1M tokens | $6.25 / 1M tokens | $17 / 1M tokens | - | N/A |
| Anthropic | Opus 4.7 | $5 / 1M tokens | $0.50 / 1M tokens | $6.25 / 1M tokens | $17 / 1M tokens | - | N/A |
| Anthropic | Opus 4.6 | $5 / 1M tokens | $0.50 / 1M tokens | $6.25 / 1M tokens | $17 / 1M tokens | - | N/A |
| OpenAI | Web search (gpt-5.4-mini) | $1.25 / 1M tokens | $0.13 / 1M tokens | - | $7.50 / 1M tokens | $10 / 1K searches | N/A |
Average costs are based on recent usage data. N/A means that no benchmark has been measured yet.
Bulk models​
Bulk models handle bulk tasks and web searches. You use them with bulk tools and GPT functions, and the Agent uses them when it runs bulk tasks or web searches.
- Subscription
- Pay-as-you-go
Built-in models​
| Provider | Model | Input | Cached input | Cache write | Output | Request | Average cost (spreadsheets) |
|---|---|---|---|---|---|---|---|
| OpenAI | gpt-5.5 | $7.14 / 1M tokens | $0.72 / 1M tokens | - | $42.86 / 1M tokens | $1 / 1K requests | N/A |
| OpenAI | gpt-5.4 | $7.14 / 1M tokens | $0.72 / 1M tokens | - | $42.86 / 1M tokens | $1 / 1K requests | N/A |
| OpenAI | gpt-5.4-mini | $1.07 / 1M tokens | $0.11 / 1M tokens | - | $6.43 / 1M tokens | $1 / 1K requests | N/A |
| OpenAI | gpt-5.4-nano | $0.29 / 1M tokens | $0.03 / 1M tokens | - | $1.79 / 1M tokens | $1 / 1K requests | N/A |
| OpenAI | gpt-5.2 | $5 / 1M tokens | $0.50 / 1M tokens | - | $40 / 1M tokens | $1 / 1K requests | N/A |
| OpenAI | gpt-5.1 | $3.57 / 1M tokens | $0.36 / 1M tokens | - | $28.57 / 1M tokens | $1 / 1K requests | N/A |
| OpenAI | gpt-5-mini | $0.64 / 1M tokens | $0.07 / 1M tokens | - | $5.14 / 1M tokens | $1 / 1K requests | N/A |
| OpenAI | gpt-5-nano | $0.07 / 1M tokens | $0.01 / 1M tokens | - | $0.57 / 1M tokens | $1 / 1K requests | N/A |
| OpenAI | gpt-4.1 | $5 / 1M tokens | $1.25 / 1M tokens | - | $20 / 1M tokens | $1 / 1K requests | N/A |
| OpenAI | gpt-4.1-mini | $1 / 1M tokens | $0.25 / 1M tokens | - | $4 / 1M tokens | $1 / 1K requests | N/A |
| OpenAI | gpt-4.1-nano | $0.14 / 1M tokens | $0.04 / 1M tokens | - | $0.57 / 1M tokens | $1 / 1K requests | N/A |
| OpenAI | gpt-5-search-api | $1.79 / 1M tokens | $0.18 / 1M tokens | - | $14.29 / 1M tokens | $14.29 / 1K requests | N/A |
| OpenAI | gpt-4o | $6.07 / 1M tokens | $3.04 / 1M tokens | - | $24.29 / 1M tokens | $1 / 1K requests | N/A |
| OpenAI | gpt-4o-mini | $0.36 / 1M tokens | $0.18 / 1M tokens | - | $1.43 / 1M tokens | $1 / 1K requests | N/A |
| OpenAI | gpt-5 | $3.57 / 1M tokens | $0.36 / 1M tokens | - | $28.57 / 1M tokens | $1 / 1K requests | N/A |
| OpenAI | gpt-4-turbo | $14.29 / 1M tokens | - | - | $42.86 / 1M tokens | $1 / 1K requests | N/A |
| OpenAI | o1 | $21.43 / 1M tokens | - | - | $85.72 / 1M tokens | $1 / 1K requests | N/A |
| OpenAI | o3-mini | $1.57 / 1M tokens | $0.79 / 1M tokens | - | $6.29 / 1M tokens | $1 / 1K requests | N/A |
| OpenAI | o3 | $2.86 / 1M tokens | $0.72 / 1M tokens | - | $11.43 / 1M tokens | $1 / 1K requests | N/A |
| OpenAI | o4-mini | $1.57 / 1M tokens | $0.39 / 1M tokens | - | $6.29 / 1M tokens | $1 / 1K requests | N/A |
| OpenAI | gpt-3.5-turbo | $0.72 / 1M tokens | - | - | $2.14 / 1M tokens | $1 / 1K requests | N/A |
 Perplexity | sonar | $1.43 / 1M tokens | - | - | $1.43 / 1M tokens | $7.14, $11.43, or $17.14 / 1K requests | N/A |
| Perplexity | sonar-pro | $4.29 / 1M tokens | - | - | $21.43 / 1M tokens | $8.57, $14.29, or $20 / 1K requests | N/A |
 Google | gemini-2.5-flash | $0.43 / 1M tokens | - | - | $3.57 / 1M tokens | $50 / 1K requests | N/A |
| Anthropic | claude-5-sonnet | $2.86 / 1M tokens | $0.29 / 1M tokens | $3.57 / 1M tokens | $14.29 / 1M tokens | $1 / 1K requests | N/A |
| Anthropic | claude-4.6-sonnet | $4.29 / 1M tokens | $0.43 / 1M tokens | $5.36 / 1M tokens | $21.43 / 1M tokens | $1 / 1K requests | N/A |
| Anthropic | claude-4.5-sonnet | $4.29 / 1M tokens | $0.43 / 1M tokens | $5.36 / 1M tokens | $21.43 / 1M tokens | $1 / 1K requests | N/A |
| Anthropic | claude-4.6-opus | $7.14 / 1M tokens | $0.72 / 1M tokens | $8.93 / 1M tokens | $35.72 / 1M tokens | $1 / 1K requests | N/A |
| Anthropic | claude-4.7-opus | $7.14 / 1M tokens | $0.72 / 1M tokens | $8.93 / 1M tokens | $35.72 / 1M tokens | $1 / 1K requests | N/A |
| Anthropic | claude-4.5-opus | $7.14 / 1M tokens | $0.72 / 1M tokens | $8.93 / 1M tokens | $35.72 / 1M tokens | $1 / 1K requests | N/A |
| Anthropic | claude-4.5-haiku | $1.43 / 1M tokens | $0.14 / 1M tokens | - | $7.14 / 1M tokens | $1 / 1K requests | N/A |
 Parallel | speed | - | - | - | - | $7.14 / 1K requests | N/A |
Average costs are based on recent usage data. N/A means that no benchmark has been measured yet.
Provider-connected models​
| Provider | Model | Input | Cached input | Cache write | Output | Request | Average cost (spreadsheets) |
|---|---|---|---|---|---|---|---|
| Varies | All provider-connected models | - | - | - | - | $1 / 1K requests | N/A |
Average costs are based on recent usage data. N/A means that no benchmark has been measured yet.
Built-in models​
| Provider | Model | Input | Cached input | Cache write | Output | Request | Average cost (spreadsheets) |
|---|---|---|---|---|---|---|---|
| OpenAI | gpt-5.5 | $50 / 1M tokens | $12.50 / 1M tokens | - | $50 / 1M tokens | - | N/A |
| OpenAI | gpt-5.4 | $50 / 1M tokens | $12.50 / 1M tokens | - | $50 / 1M tokens | - | $25.27 / 1K rows |
| OpenAI | gpt-5.4-mini | $9 / 1M tokens | $2.25 / 1M tokens | - | $9 / 1M tokens | - | $3.41 / 1K rows |
| OpenAI | gpt-5.4-nano | $2.50 / 1M tokens | $0.63 / 1M tokens | - | $2.50 / 1M tokens | - | $1.65 / 1K rows |
| OpenAI | gpt-5.2 | $50 / 1M tokens | $12.50 / 1M tokens | - | $50 / 1M tokens | - | $32.21 / 1K rows |
| OpenAI | gpt-5.1 | $40 / 1M tokens | $10 / 1M tokens | - | $40 / 1M tokens | - | $23.47 / 1K rows |
| OpenAI | gpt-5-mini | $8 / 1M tokens | $2 / 1M tokens | - | $8 / 1M tokens | - | $6.75 / 1K rows |
| OpenAI | gpt-5-nano | $1 / 1M tokens | $0.25 / 1M tokens | - | $1 / 1M tokens | - | $1.18 / 1K rows |
| OpenAI | gpt-4.1 | $30 / 1M tokens | $7.50 / 1M tokens | - | $30 / 1M tokens | - | $11.23 / 1K rows |
| OpenAI | gpt-4.1-mini | $6 / 1M tokens | $1.50 / 1M tokens | - | $6 / 1M tokens | - | $2.36 / 1K rows |
| OpenAI | gpt-4.1-nano | $1 / 1M tokens | $0.25 / 1M tokens | - | $1 / 1M tokens | - | $0.45 / 1K rows |
| OpenAI | gpt-5-search-api | $3 / 1M tokens | $3 / 1M tokens | - | $13 / 1M tokens | $15 / 1K searches | $62.03 / 1K rows |
| OpenAI | gpt-4o | $30 / 1M tokens | $7.50 / 1M tokens | - | $30 / 1M tokens | - | $7.48 / 1K rows |
| OpenAI | gpt-4o-mini | $2 / 1M tokens | $0.50 / 1M tokens | - | $2 / 1M tokens | - | $0.98 / 1K rows |
| OpenAI | gpt-5 | $40 / 1M tokens | $10 / 1M tokens | - | $40 / 1M tokens | - | $27.41 / 1K rows |
| OpenAI | gpt-4-turbo | $35 / 1M tokens | - | - | $35 / 1M tokens | - | $24.89 / 1K rows |
| OpenAI | o1 | $70 / 1M tokens | $17.50 / 1M tokens | - | $70 / 1M tokens | - | $26.81 / 1K rows |
| OpenAI | o3-mini | $10 / 1M tokens | $2.50 / 1M tokens | - | $10 / 1M tokens | - | N/A |
| OpenAI | o3 | $10 / 1M tokens | $2.50 / 1M tokens | - | $10 / 1M tokens | - | $6.48 / 1K rows |
| OpenAI | o4-mini | $10 / 1M tokens | $2.50 / 1M tokens | - | $10 / 1M tokens | - | $4.03 / 1K rows |
| OpenAI | gpt-3.5-turbo | $2.50 / 1M tokens | - | - | $2.50 / 1M tokens | - | $1.46 / 1K rows |
| Perplexity | sonar | $10 / 1M tokens | - | - | $10 / 1M tokens | $10, $13, or $17 / 1K searches | $8.88 / 1K rows |
| Perplexity | sonar-pro | $20 / 1M tokens | - | - | $20 / 1M tokens | $13, $15, or $19 / 1K searches | $15.78 / 1K rows |
| Google | gemini-2.5-flash | $4 / 1M tokens | $1 / 1M tokens | - | $4 / 1M tokens | $40 / 1K searches | $15.56 / 1K rows |
| Anthropic | claude-5-sonnet | $30 / 1M tokens | $7.50 / 1M tokens | $33.75 / 1M tokens | $30 / 1M tokens | - | N/A |
| Anthropic | claude-4.6-sonnet | $30 / 1M tokens | $7.50 / 1M tokens | $33.75 / 1M tokens | $30 / 1M tokens | - | $16.12 / 1K rows |
| Anthropic | claude-4.5-sonnet | $20 / 1M tokens | $5 / 1M tokens | $23.75 / 1M tokens | $20 / 1M tokens | - | $12.86 / 1K rows |
| Anthropic | claude-4.6-opus | $50 / 1M tokens | $12.50 / 1M tokens | $56.25 / 1M tokens | $50 / 1M tokens | - | $10.26 / 1K rows |
| Anthropic | claude-4.7-opus | $50 / 1M tokens | $12.50 / 1M tokens | $56.25 / 1M tokens | $50 / 1M tokens | - | N/A |
| Anthropic | claude-4.5-opus | $50 / 1M tokens | $12.50 / 1M tokens | $56.25 / 1M tokens | $50 / 1M tokens | - | $179.35 / 1K rows |
| Anthropic | claude-4.5-haiku | $6 / 1M tokens | $1.50 / 1M tokens | - | $6 / 1M tokens | - | $3.99 / 1K rows |
| Parallel | speed | - | - | - | - | $10 / 1K searches | $7 / 1K rows |
Average costs are based on recent usage data. N/A means that no benchmark has been measured yet.
Provider-connected models​
| Provider | Model | Input | Cached input | Cache write | Output | Request | Average cost (spreadsheets) |
|---|---|---|---|---|---|---|---|
| OpenAI | gpt-4o-mini | $0.30 / 1M tokens | - | - | $0.30 / 1M tokens | - | N/A |
| OpenAI | gpt-4o-mini (fine-tuned) | $0.30 / 1M tokens | - | - | $0.30 / 1M tokens | - | N/A |
| Varies | All other models | $1 / 1M tokens | - | - | $1 / 1M tokens | - | N/A |
Average costs are based on recent usage data. N/A means that no benchmark has been measured yet.