API models and pricing
One key to use TuttiFrutti's models from your code, your agent or your product. It speaks the OpenAI format
(/v1/chat/completions) and the Claude format (/v1/messages), so it works with their SDKs and
with any compatible tool.
Models
Price per million tokens, in US dollars. Paid with credits (1,000 credits = $1): your plan's or a top-up's, with no minimums and no per-key fees.
| Model | Input | Cached input | Output | Plan |
|---|---|---|---|---|
TuttiFruttituttifrutti | … | … | … | … |
TuttiFrutti Protuttifrutti-pro | … | … | … | … |
DeepSeek Flashdeepseek-flash | … | … | … | … |
DeepSeek V4 Prodeepseek-v4-pro | … | … | … | … |
| Images, voice and video | Path | Price | Plan |
|---|
Start in a minute
OpenAI format
curl https://tuttifrutti.app/v1/chat/completions \
-H "Authorization: Bearer $TUTTIFRUTTI_API_KEY" \
-H "Content-Type: application/json" \
-d '{"model": "tuttifrutti-pro", "messages": [{"role": "user", "content": "Hello"}]}'
With the OpenAI SDK, just set the base URL to https://tuttifrutti.app/v1 and use your key.
Claude format (agents that use it)
ANTHROPIC_BASE_URL=https://tuttifrutti.app
ANTHROPIC_API_KEY=chk_…
ANTHROPIC_MODEL=tuttifrutti-pro
ANTHROPIC_SMALL_FAST_MODEL=tuttifrutti
The key can go in x-api-key or as a Bearer token. Text, images, tools and streamed replies are supported.
How you're charged
- Each reply is charged for the tokens it uses: input, input already cached (much cheaper) and output.
- If a provider doesn't answer, the request fails and isn't charged. The API never uses the chat's free messages.
- Credits are bought in euros, VAT included. Top-up credits never expire.
- Your plan decides which models you can use and how many requests at once.