> ## Documentation Index > Fetch the complete documentation index at: https://docs.omniall.ai/llms.txt > Use this file to discover all available pages before exploring further. # Gemini streamGenerateContent > Google Gemini `streamGenerateContent` relayed by Omniall AI. **Path:** `POST https://api.omniall.ai/v1beta/models/{model}:streamGenerateContent` Streaming native Gemini call (SSE). Auth: `Authorization: Bearer sk-...`, or `x-goog-api-key` / `?key=`. ### Scenarios - **Vision**: `inlineData` image parts - **Files**: `fileData.fileUri` - **Grounding / web**: `tools: [{ googleSearch: {} }]` when supported - **Agent**: `functionDeclarations` - **Image generation**: `generationConfig.responseModalities` including `IMAGE` Official reference: https://ai.google.dev/api/generate-content ## OpenAPI ````yaml openapi/xgapi-public.yaml POST /v1beta/models/{model}:streamGenerateContent openapi: 3.0.3 info: title: Omniall AI Public API description: >- Omniall AI public API. Website: https://omniall.ai — API host: https://api.omniall.ai — Model Square: https://omniall.ai/pricing version: 1.1.0 servers: - url: https://api.omniall.ai description: Production security: - bearerAuth: [] tags: - name: Models description: List models available to your API key. - name: ChatGPT Chat Completions x-group: ChatGPT description: OpenAI Chat Completions (`POST /v1/chat/completions`). - name: Responses description: OpenAI Responses API (`POST /v1/responses`). - name: Claude description: Anthropic Messages API (`POST /v1/messages`). - name: Gemini description: Google Gemini generateContent / streamGenerateContent. - name: Images description: >- Image generation (OpenAI Images), vendor guides, and Midjourney proxy APIs. - name: Video description: Async video tasks. - name: Audio description: Speech and related audio APIs. - name: Rerank description: Document reranking. paths: /v1beta/models/{model}:streamGenerateContent: post: tags: - Gemini summary: Gemini streamGenerateContent description: > Google Gemini `streamGenerateContent` relayed by Omniall AI. **Path:** `POST https://api.omniall.ai/v1beta/models/{model}:streamGenerateContent` Streaming native Gemini call (SSE). Auth: `Authorization: Bearer sk-...`, or `x-goog-api-key` / `?key=`. ### Scenarios - **Vision**: `inlineData` image parts - **Files**: `fileData.fileUri` - **Grounding / web**: `tools: [{ googleSearch: {} }]` when supported - **Agent**: `functionDeclarations` - **Image generation**: `generationConfig.responseModalities` including `IMAGE` Official reference: https://ai.google.dev/api/generate-content operationId: geminiStreamGenerateContent parameters: - name: model in: path required: true schema: type: string example: gemini-2.0-flash description: Gemini model id (as exposed in Model Square / GET /v1/models). - name: key in: query required: false schema: type: string description: Optional Gemini-style API key query param. - name: x-goog-api-key in: header required: false schema: type: string description: Optional Gemini-style API key header. requestBody: required: true content: application/json: schema: $ref: '#/components/schemas/GeminiGenerateContentRequest' examples: basic: summary: Basic generateContent value: contents: - role: user parts: - text: Hello Gemini vision: summary: Image understanding value: contents: - role: user parts: - text: Caption this image - inlineData: mimeType: image/jpeg data: /9j/4AAQ... file: summary: File URI analysis value: contents: - role: user parts: - text: Summarize the PDF - fileData: mimeType: application/pdf fileUri: https://example.com/doc.pdf google_search: summary: Google Search grounding value: contents: - role: user parts: - text: Who won the latest F1 race? tools: - googleSearch: {} function_calling: summary: Function calling (agent) value: contents: - role: user parts: - text: Weather in Paris? tools: - functionDeclarations: - name: get_weather description: Get weather parameters: type: object properties: city: type: string required: - city responses: '200': description: Gemini response JSON (or SSE for streamGenerateContent). content: application/json: schema: $ref: '#/components/schemas/GeminiGenerateContentResponse' '400': description: Invalid request body or unsupported parameter for the model. content: application/json: schema: $ref: '#/components/schemas/OpenAIErrorResponse' '401': description: Missing or invalid API key. content: application/json: schema: $ref: '#/components/schemas/OpenAIErrorResponse' example: error: message: '无效的令牌 (request id: ...)' type: new_api_error code: '' '403': description: Forbidden (banned user, IP allowlist, group access). content: application/json: schema: $ref: '#/components/schemas/OpenAIErrorResponse' '429': description: Rate limited. content: application/json: schema: $ref: '#/components/schemas/OpenAIErrorResponse' '500': description: Upstream or gateway internal error. content: application/json: schema: $ref: '#/components/schemas/OpenAIErrorResponse' components: schemas: GeminiGenerateContentRequest: type: object required: - contents description: >- Native Gemini `generateContent` request body, relayed by Omniall AI. One `contents` turn is the minimum; add prior user/model turns for multi-turn chat. properties: contents: type: array items: type: object properties: role: type: string enum: - user - model description: >- Who produced this turn. `user` = your application / end user input; `model` = previous Gemini assistant output (include these when continuing a multi-turn chat so the model has context). parts: type: array items: type: object properties: text: type: string description: >- Plain text for this part (prompt, question, or prior model reply). Example: "Describe this image". example: Describe this image inlineData: type: object properties: mimeType: type: string example: image/jpeg description: >- IANA MIME type of the binary data. Common values: `image/jpeg`, `image/png`, `image/webp`, `application/pdf`, `audio/mpeg`. Must match the real file type or the model may reject the part. data: type: string description: >- Raw file bytes encoded as Base64 **without** a `data:` URL prefix. Do not include `data:image/jpeg;base64,` here—only the Base64 payload. Large files increase request size and latency; keep images reasonably sized. example: /9j/4AAQSkZJRgABAQ... description: >- Inline binary media (image, audio, PDF, etc.) encoded as Base64. Prefer this when you have file bytes available and no public URL. fileData: type: object properties: mimeType: type: string description: >- MIME type of the remote file (same rules as `inlineData.mimeType`). example: application/pdf fileUri: type: string description: >- Publicly reachable file URI/URL. Must be downloadable without login cookies. Prefer HTTPS. example: https://example.com/report.pdf description: >- Reference to a remote file instead of inlining Base64. Use when the file is already reachable by URL / URI that the upstream accepts. description: >- One or more content parts for this turn. A turn can mix text and media (for example text + an image). At least one part is required. description: A single conversation turn. description: >- Ordered conversation turns sent to Gemini. Each item is one message from either the user or the model. For a first request, usually a single `user` turn is enough. systemInstruction: type: object properties: parts: type: array items: type: object additionalProperties: true description: System instruction parts (usually one text part). description: >- Optional system-level instruction that steers model behavior for the whole request (persona, output format, safety rules). Not counted as a user/model turn in `contents`. tools: type: array description: >- Optional tools the model may call. Examples: `{ "googleSearch": {} }` for grounding/web search (when enabled), or `{ "functionDeclarations": [ ... ] }` for your own agent functions. Availability depends on the model/channel. items: type: object additionalProperties: true toolConfig: type: object additionalProperties: true description: >- Controls how tools are selected (for example function-calling mode). Pass through as Gemini `toolConfig` when you need to force/auto function calls. generationConfig: type: object properties: temperature: type: number description: >- Randomness of the answer. Higher = more creative/varied; lower = more deterministic. Typical range about 0–2 depending on model. topP: type: number description: >- Nucleus sampling: only tokens within the top probability mass `p` are considered. Lower values make output more focused. topK: type: integer description: >- Only sample from the top-K tokens. Smaller K = more conservative wording. maxOutputTokens: type: integer description: >- Hard cap on how many tokens the model may generate in the response. Increase for long answers; decrease to control cost/latency. responseMimeType: type: string description: >- Desired MIME type of the model text response, e.g. `text/plain` or `application/json` for JSON-oriented outputs when supported. responseModalities: type: array items: type: string description: >- Requested response modalities. For image generation include `IMAGE`, e.g. `["TEXT","IMAGE"]` or `["IMAGE"]`. example: - TEXT - IMAGE thinkingConfig: type: object additionalProperties: true description: >- Optional thinking/reasoning controls for models that support them (channel/model dependent). imageConfig: type: object description: >- Image output settings when `responseModalities` includes `IMAGE`. Snake_case aliases are also accepted. properties: aspectRatio: type: string description: >- Output aspect ratio, e.g. `1:1`, `16:9`, `9:16`, `3:4`, `4:3`. enum: - '1:1' - '2:3' - '3:2' - '3:4' - '4:3' - '4:5' - '5:4' - '9:16' - '16:9' - '21:9' example: '16:9' imageSize: type: string description: 'Resolution tier: `1K` / `2K` / `4K`.' enum: - 1K - 2K - 4K example: 2K description: Sampling and output controls for this generation. safetySettings: type: array items: type: object additionalProperties: true description: >- Per-category safety thresholds (harassment, hate, sexual, dangerous content, etc.). Use when you need stricter or looser filtering than defaults. GeminiGenerateContentResponse: type: object properties: candidates: type: array items: type: object properties: content: type: object properties: role: type: string description: Usually `model`. parts: type: array items: type: object additionalProperties: true description: >- Answer parts (`text`, inline images, function calls, etc.). description: Model message content for this candidate. finishReason: type: string description: >- Why generation stopped (e.g. STOP, MAX_TOKENS, SAFETY). Check this if the answer looks truncated or empty. description: A single candidate. description: One or more candidate replies. Usually use `candidates[0]`. usageMetadata: type: object properties: promptTokenCount: type: integer description: Tokens in the request (contents + system + tools, etc.). candidatesTokenCount: type: integer description: Tokens generated in the response. totalTokenCount: type: integer description: Sum of prompt and candidates tokens. description: Token usage for billing/debugging. description: >- Gemini generateContent response. Read the first candidate’s `content.parts` for the answer text/images. OpenAIErrorResponse: type: object description: OpenAI-style error envelope returned on auth/request failures. required: - error properties: error: type: object required: - message - type properties: message: type: string description: >- Human-readable error. The gateway may append a request id for support. type: type: string description: Error type, commonly `new_api_error`. example: new_api_error code: type: string description: >- Optional machine-readable code (e.g. `access_denied`). May be an empty string. description: Error object. securitySchemes: bearerAuth: type: http scheme: bearer bearerFormat: API Key description: 'Use Authorization: Bearer sk-... from https://omniall.ai/dashboard' ```` This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.