> ## Documentation Index > Fetch the complete documentation index at: https://docs.shaktistudio.ai/llms.txt > Use this file to discover all available pages before exploring further. # Gemma 3 4B > Create a chat completion for given messages with streaming support using Google's Gemma-IT model, optimized for instruction following and conversational AI tasks ## OpenAPI ````yaml /openapi-specs/playground/gemma3-4b.yaml POST /chat/completions openapi: 3.1.0 info: title: Gemma3 4B description: >- Google's Gemma-IT Language Model API for conversational AI, text generation, and instruction-following tasks with built-in retry mechanisms version: 1.0.0 servers: - url: https://http.llm.model-cluster.on-prem.clusters.yotta-uat.cluster.s9t.link description: Production LLM Proxy Server security: - BearerAuth: [] tags: - name: Chat Completion description: Gemma-IT language model chat completion services paths: /chat/completions: post: tags: - Chat Completion summary: Create chat completion with Gemma-IT description: >- Create a chat completion for given messages with streaming support using Google's Gemma-IT model, optimized for instruction following and conversational AI tasks operationId: createChatCompletionGemmaIT parameters: - in: header name: id schema: type: string default: 54e4c0bb-11fc-40f6-94d0-5a6eca081345 required: true description: Model UUID for the request (Gemma-IT model identifier) example: 54e4c0bb-11fc-40f6-94d0-5a6eca081345 requestBody: required: true content: application/json: schema: type: object properties: messages: type: array description: Array of messages in the conversation items: type: object properties: role: type: string enum: - system - user - assistant description: The role of the message sender content: type: string description: The content of the message required: - role - content model: type: string description: Model identifier for Gemma-IT enum: - gemma-it default: gemma-it stream: type: boolean description: Whether to stream the response default: false temperature: type: number description: >- Sampling temperature (0.0 to 2.0). Controls randomness in generation minimum: 0 maximum: 2 default: 0.7 max_tokens: type: integer description: Maximum number of tokens to generate minimum: 1 maximum: 4096 default: 1024 top_p: type: number description: >- Nucleus sampling parameter for controlling response diversity minimum: 0 maximum: 1 default: 0.9 top_k: type: integer description: Top-k sampling parameter for vocabulary selection minimum: 1 maximum: 100 default: 40 stop: type: array description: Sequences where the API will stop generating further tokens items: type: string maxItems: 4 nullable: true frequency_penalty: type: number description: Penalty for frequent tokens to reduce repetition minimum: -2 maximum: 2 default: 0 presence_penalty: type: number description: Penalty for new tokens to encourage topic diversity minimum: -2 maximum: 2 default: 0 repetition_penalty: type: number description: Penalty for repeating tokens (Gemma-specific parameter) minimum: 0.1 maximum: 2 default: 1 do_sample: type: boolean description: Whether to use sampling for generation default: true seed: type: integer description: Random seed for reproducible outputs nullable: true system_prompt: type: string description: >- System prompt to set model behavior (alternative to system message) nullable: true required: - messages - model examples: simple_math_query: summary: Simple arithmetic question value: model: gemma-it messages: - role: user content: what is 2+2? stream: false max_tokens: 1024 instruction_following: summary: Instruction following task value: model: gemma-it messages: - role: system content: >- You are a helpful assistant that follows instructions precisely. - role: user content: >- Write a brief summary of the benefits of renewable energy in exactly 3 bullet points. stream: false temperature: 0.3 max_tokens: 512 top_p: 0.95 conversational_chat: summary: Conversational interaction value: model: gemma-it messages: - role: user content: >- Can you explain how machine learning works in simple terms? stream: false temperature: 0.8 max_tokens: 2048 top_p: 0.9 top_k: 50 creative_writing: summary: Creative content generation value: model: gemma-it messages: - role: user content: Write a short poem about technology and human connection stream: false temperature: 1 max_tokens: 800 top_k: 60 multi_turn_conversation: summary: Multi-turn conversation with context value: model: gemma-it messages: - role: user content: What's the capital of Japan? - role: assistant content: The capital of Japan is Tokyo. - role: user content: What's the population of that city? stream: true temperature: 0.5 max_tokens: 512 responses: '200': description: Successful chat completion content: application/json: schema: oneOf: - $ref: '#/components/schemas/ChatCompletionResponse' - $ref: '#/components/schemas/ChatCompletionChunk' text/event-stream: schema: type: string description: Server-sent events for streaming responses example: > data: {"id":"chatcmpl-gemma-it-abc123","object":"chat.completion.chunk","created":1699014493,"model":"gemma-it","choices":[{"index":0,"delta":{"content":"2"},"finish_reason":null}]} data: {"id":"chatcmpl-gemma-it-abc123","object":"chat.completion.chunk","created":1699014493,"model":"gemma-it","choices":[{"index":0,"delta":{"content":" + 2"},"finish_reason":null}]} data: {"id":"chatcmpl-gemma-it-abc123","object":"chat.completion.chunk","created":1699014493,"model":"gemma-it","choices":[{"index":0,"delta":{"content":" = 4"},"finish_reason":"stop"}]} '400': description: Bad Request - Invalid request parameters content: application/json: schema: $ref: '#/components/schemas/Error' examples: invalid_model: summary: Invalid model specified value: error: message: Invalid model specified. Expected 'gemma-it' type: invalid_request_error code: invalid_model token_limit_exceeded: summary: Token limit exceeded value: error: message: max_tokens cannot exceed 4096 type: invalid_request_error code: invalid_parameter param: max_tokens empty_messages: summary: Empty messages array value: error: message: messages array cannot be empty type: invalid_request_error code: invalid_parameter param: messages '401': description: Unauthorized - Invalid or missing authentication content: application/json: schema: $ref: '#/components/schemas/Error' '429': description: Too Many Requests - Rate limit exceeded content: application/json: schema: $ref: '#/components/schemas/Error' '500': description: Internal Server Error content: application/json: schema: $ref: '#/components/schemas/Error' '502': description: Bad Gateway - Gemma model service unavailable content: application/json: schema: $ref: '#/components/schemas/Error' '503': description: Service Unavailable - Model temporarily overloaded content: application/json: schema: $ref: '#/components/schemas/Error' components: schemas: ChatCompletionResponse: type: object description: Response for non-streaming chat completion properties: id: type: string description: Unique identifier for the completion example: chatcmpl-gemma-it-abc123 object: type: string enum: - chat.completion description: Object type example: chat.completion created: type: integer description: Unix timestamp of when the completion was created example: 1699014493 model: type: string description: The model used for completion example: gemma-it choices: type: array items: type: object properties: index: type: integer description: Index of the choice example: 0 message: type: object properties: role: type: string enum: - assistant description: Role of the message sender example: assistant content: type: string description: Generated text content example: >- 2 + 2 equals 4. This is a basic arithmetic operation where we add two numbers together to get their sum. required: - role - content finish_reason: type: string enum: - stop - length - content_filter - tool_calls description: Reason for finishing the generation example: stop required: - index - message - finish_reason usage: type: object properties: prompt_tokens: type: integer description: Number of tokens in the prompt example: 8 completion_tokens: type: integer description: Number of tokens in the completion example: 24 total_tokens: type: integer description: Total number of tokens used example: 32 required: - prompt_tokens - completion_tokens - total_tokens system_fingerprint: type: string description: System fingerprint for the model version example: gemma-2-instruct-v1.0 nullable: true ChatCompletionChunk: type: object description: Response chunk for streaming chat completion properties: id: type: string description: Unique identifier for the completion example: chatcmpl-gemma-it-abc123 object: type: string enum: - chat.completion.chunk description: Object type example: chat.completion.chunk created: type: integer description: Unix timestamp of when the completion was created example: 1699014493 model: type: string description: The model used for completion example: gemma-it choices: type: array items: type: object properties: index: type: integer description: Index of the choice example: 0 delta: type: object properties: role: type: string enum: - assistant description: Role of the message sender (only in first chunk) example: assistant content: type: string description: Generated text content chunk nullable: true example: ' + 2' finish_reason: type: string enum: - stop - length - content_filter - tool_calls description: Reason for finishing the generation (only in last chunk) nullable: true example: null required: - index - delta system_fingerprint: type: string description: System fingerprint for the model version example: gemma-2-instruct-v1.0 nullable: true Error: type: object properties: error: type: object properties: message: type: string description: Human-readable error message example: Invalid model specified type: type: string description: Error type classification example: invalid_request_error enum: - invalid_request_error - authentication_error - rate_limit_error - server_error - model_error code: type: string description: Specific error code example: invalid_model param: type: string description: Parameter that caused the error (if applicable) example: model nullable: true required: - message - type securitySchemes: BearerAuth: type: http scheme: bearer bearerFormat: JWT description: JWT token for authentication - Use your API token as the Bearer token ````