Text Generation (Gemini)

POSThttps://api.tokenbyte.ai/v1beta/models/{model}:generateContent

Use streamGenerateContent for streaming. Authenticate with x-goog-api-key: <TOKENBYTE_API_KEY>.

Request body

Field	Type	Required	Description
`contents`	array	Yes	Conversation turns containing `role` and `parts`
`systemInstruction`	object	No	System instruction
`generationConfig`	object	No	Sampling, output limit, and stop settings
`tools`, `toolConfig`	array/object	No	Tools and calling policy
`safetySettings`	array	No	Safety thresholds when supported

Parts may contain text, inlineData, fileData, functionCall, or functionResponse. Read output from candidates[].content.parts; token usage is returned in usageMetadata.

curl "https://api.tokenbyte.ai/v1beta/models/<model-id>:generateContent" \
  -H "x-goog-api-key: $TOKENBYTE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"contents":[{"role":"user","parts":[{"text":"Hello"}]}]}'

{
  "candidates": [
    {
      "content": { "role": "model", "parts": [{ "text": "Hello!" }] },
      "finishReason": "STOP"
    }
  ],
  "usageMetadata": {
    "promptTokenCount": 8,
    "candidatesTokenCount": 4,
    "totalTokenCount": 12
  }
}