Pricing
WritersAPI has no plans, seats or subscriptions. Pick how each site pays for the AI it uses: through our provider accounts at cost plus 5%, or with its own keys for nothing.
No subscription. No minimum. No card on file.+5% on provider cost
Use WritersAPI's own accounts with every provider. Nothing to set up.
$0 forever
Add your DeepSeek, Gemini, OpenRouter or FLUX keys. Your providers bill you directly.
Provider list prices plus 5%, in US dollars. Image models priced by tokens bill the tokens the image uses (about 1,300 per image). Where a provider reports the exact cost of a call, that figure is used instead.
| Model id | Family | Task | Status | Input / 1M | Output / 1M | Per image |
|---|---|---|---|---|---|---|
| gemini-3.8-flash | gemini | Writing | With your own key | $0.79 | $3.94 | n/a |
| gpt-6.1-sol | chatgpt | Writing | With your own key | $2.10 | $10.50 | n/a |
| deepseek-v4-pro | deepseek | Writing | With your own key | $1.39 | $4.16 | n/a |
| gemini-4-argon | gemini | Writing | With your own key | $2.10 | $10.50 | n/a |
| gpt-6-astra | chatgpt | Writing | With your own key | $10.50 | $52.50 | n/a |
| deepseek-flash | deepseek | Writing | With your own key | $0.32 | $1.26 | n/a |
| gpt-6-luna | chatgpt | Writing | With your own key | $0.10 | $0.53 | n/a |
| gemini-3.1-flash-lite | gemini | Writing | With your own key | $0.26 | $1.57 | n/a |
| flux-2-pro | flux | Image | With your own key | n/a | n/a | $0.032 |
| nano-banana-2-lite | nano-banana | Image | With your own key | $0.26 | $31.50 | by tokens |
| gpt-image-2 | chatgpt | Image | With your own key | $8.40 | $31.50 | by tokens |
| flux-3 | flux | Image | With your own key | n/a | n/a | $0.063 |
| nano-banana-2.1 | nano-banana | Image | With your own key | $1.57 | $31.50 | by tokens |
| gpt-image-mini | chatgpt | Image | With your own key | $2.63 | $8.40 | by tokens |
| flux-2-klein | flux | Image | With your own key | n/a | n/a | $0.015 |
WritersAPI is maintained by PRNow as a free tool for publishers. The 5% on hosted keys covers running costs; with your own keys there is nothing to cover.
Input and output tokens (including thinking tokens), images generated, and web search when you turn it on. Failed attempts that moved to another model cost nothing; only the model that answered is billed.
Yes. Each API key can have a monthly budget. When it is reached, requests answer 402 until the next month (UTC).
No. Requests only route to providers you have a key for. Add more providers for more fallback options, or allow fallback to hosted keys for the rest.