Gemini native

Call Gemini 4 Argon, Gemini 3.8 Flash and Gemini 3.1 Pro with Gemini generateContent requests and one BeatAPI key.

• POST /v1beta/models/{model}:generateContent, and :streamGenerateContent?alt=sse for streaming, for every current Gemini model.
• Text and image input, text output, streaming, function calling and thinking.
• Send the key in the x-goog-api-key header or as a Bearer token.
• The model is in the URL, not in the body.

Opens this endpoint in the API Explorer. Enter your BeatAPI API key and send: the call is real and is billed to your own balance.

Models

Put one of these IDs in the URL, in place of {model}.

• Prices are the base tier, discount already applied. A long-context tier, cached-input rates and cache writes are not in this table; see the price table.
• — means the catalogue does not publish the number.

Older IDs of this family still work and are listed on Legacy models.

Quick start

Keep the key in BEATAPI_API_KEY and never put it in browser code. The model ID goes in the URL, so the body has no model field. These are request examples, not recorded model outputs.

Read the answer from candidates[0].content.parts[0].text, then check candidates[0].finishReason and usageMetadata. An HTTP 200 alone does not prove the answer satisfies your task: validate it in your application. Text requests return directly; do not poll /v1/tasks for them.

curl --request POST \
  --url https://direct.beatapi.io/v1beta/models/gemini-3.8-flash:generateContent \
  --header "x-goog-api-key: $BEATAPI_API_KEY" \
  --header 'Content-Type: application/json' \
  --data '{
  "contents": [
    {
      "role": "user",
      "parts": [
        {
          "text": "Design a resilient webhook retry strategy for a payments API."
        }
      ]
    }
  ]
}'

Endpoint and authentication

POST https://direct.beatapi.io/v1beta/models/{model}:generateContent

For streaming, call POST /v1beta/models/{model}:streamGenerateContent?alt=sse. The key goes in the x-goog-api-key header or in Authorization: Bearer <key>; both work.

:countTokens is not available and returns 404.

Request parameters

These are the fields that take effect. Other Gemini fields are accepted in the request but have no effect: generationConfig.thinkingConfig, generationConfig.responseMimeType, generationConfig.responseSchema, generationConfig.seed, safetySettings, toolConfig and cachedContent. Do not rely on them.

[
  {
    "role": "user",
    "parts": [
      { "text": "What is in this image?" },
      { "inlineData": { "mimeType": "image/png", "data": "<base64>" } }
    ]
  }
]

{
  "parts": [{ "text": "Answer in one short paragraph." }]
}

{
  "temperature": 0.7,
  "topP": 0.95,
  "maxOutputTokens": 2048,
  "stopSequences": ["END"]
}

[
  {
    "functionDeclarations": [
      {
        "name": "get_weather",
        "description": "Get the current weather for a city.",
        "parameters": {
          "type": "object",
          "properties": { "city": { "type": "string" } },
          "required": ["city"]
        }
      }
    ]
  }
]

Response

A non-streaming call returns one JSON object. This is an abbreviated example.

{
  "candidates": [
    {
      "content": {
        "role": "model",
        "parts": [
          {
            "text": "Retry with exponential backoff and jitter, and make every handler idempotent."
          }
        ]
      },
      "finishReason": "STOP",
      "index": 0
    }
  ],
  "usageMetadata": {
    "promptTokenCount": 19,
    "candidatesTokenCount": 16,
    "totalTokenCount": 120,
    "thoughtsTokenCount": 85
  }
}

Response fields

Streaming

Call :streamGenerateContent?alt=sse. Every event is a line data: {...} holding a candidates object, in the same shape as a non-streaming answer.

• The last event carries finishReason, for example "STOP".
• The response is text/event-stream. A failure before the model is called is returned as a normal JSON error with an HTTP status, not as a stream event.
• Set explicit timeouts and reconnect a streamed response safely.

data: {"candidates":[{"content":{"role":"model","parts":[{"text":"Retry"}]},"index":0}]}
data: {"candidates":[{"content":{"role":"model","parts":[{"text":" with backoff"}]},"finishReason":"STOP","index":0}],"usageMetadata":{"promptTokenCount":19,"candidatesTokenCount":16,"totalTokenCount":120,"thoughtsTokenCount":85}}

Errors

HTTP errors at a glance

Errors on the Gemini-native endpoint have one shape, {"error": {"message", "type": "beatapi_error", "code"}}: the code is in error.code, the request id ends message, and there is no retryable field. See Errors.

Full code table and retry policy: Errors.

Notes

• Gemini models also answer on Chat Completions, Responses and Anthropic Messages. See the other Gemini pages.
• Thinking tokens are billed as output tokens. Budget output for thinking plus the answer, not just the answer.
• Keep API keys on trusted servers, log the response request ID, and retry only 429, 500 and 503.