> ## Documentation Index > Fetch the complete documentation index at: https://docs.together.ai/llms.txt > Use this file to discover all available pages before exploring further. # Execute code > Executes the given code snippet and returns the output. Without a session_id, a new session is created to run the code. If you pass a valid session_id, the code runs in that session. This is useful for running multiple code snippets in the same environment, because dependencies and similar things are persisted between calls to the same session. ## OpenAPI ````yaml POST /tci/execute openapi: 3.1.0 info: title: Together APIs description: The Together REST API. See https://docs.together.ai for more details. version: 2.0.0 termsOfService: https://www.together.ai/terms-of-service contact: name: Together Support url: https://www.together.ai/contact license: name: MIT url: https://github.com/togethercomputer/openapi/blob/main/LICENSE servers: - url: https://api.together.ai/v1 security: - bearerAuth: [] paths: /tci/execute: post: tags: - Code Interpreter summary: Execute code description: > Executes the given code snippet and returns the output. Without a session_id, a new session is created to run the code. If you pass a valid session_id, the code runs in that session. This is useful for running multiple code snippets in the same environment, because dependencies and similar things are persisted between calls to the same session. operationId: tci/execute parameters: [] requestBody: content: application/json: schema: $ref: '#/components/schemas/ExecuteRequest' description: Execute Request required: false responses: '200': content: application/json: schema: $ref: '#/components/schemas/ExecuteResponse' description: Execute Response callbacks: {} x-codeSamples: - lang: Python label: Together AI SDK (v2) source: | # Docs for v1 can be found by changing the above selector ^ from together import Together import os client = Together( api_key=os.environ.get("TOGETHER_API_KEY"), ) response = client.code_interpreter.execute( code="print('Hello world!')", language="python", ) print(response.data.outputs[0].data); - lang: Python label: Together AI SDK (v1) source: | from together import Together import os client = Together( api_key=os.environ.get("TOGETHER_API_KEY"), ) response = client.code_interpreter.run( code="print('Hello world!')", language="python", ) print(response.data.outputs[0].data); - lang: TypeScript label: Together AI SDK (TypeScript) source: | import Together from "together-ai"; const client = new Together({ apiKey: process.env.TOGETHER_API_KEY, }); const response = await client.codeInterpreter.execute({ code: "print('Hello world!')", language: "python" }); console.log(response.data?.outputs?.[0]?.data); - lang: JavaScript label: Together AI SDK (JavaScript) source: | import Together from "together-ai"; const client = new Together({ apiKey: process.env.TOGETHER_API_KEY, }); const response = await client.codeInterpreter.execute({ code: "print('Hello world!')", language: "python" }); console.log(response.data?.outputs?.[0]?.data); - lang: Shell label: cURL source: | curl -X POST "https://api.together.ai/v1/tci/execute" \ -H "Authorization: Bearer $TOGETHER_API_KEY" \ -H "Content-Type: application/json" \ -d '{ "code": "print(\'Hello world!\')", "language": "python" }' components: schemas: ExecuteRequest: title: ExecuteRequest required: - language - code properties: code: description: Code snippet to execute. example: print('Hello, world!') type: string files: description: >- Files to upload to the session. If present, files are uploaded before executing the given code. items: properties: content: type: string encoding: description: >- Encoding of the file content. Use `string` for text files such as code, and `base64` for binary files, such as images. enum: - string - base64 type: string name: type: string required: - name - encoding - content type: object type: array language: default: python description: >- Programming language for the code to execute. Currently only supports Python. enum: - python session_id: description: >- Identifier of the current session. Used to make follow-up calls. Returns an error if the session does not belong to the caller or has expired. example: ses_abcDEF123 nullable: false type: string ExecuteResponse: title: ExecuteResponse type: object description: >- The result of the execution. If successful, `data` contains the result and `errors` is null. If unsuccessful, `data` is null and `errors` contains the errors. oneOf: - title: SuccessfulExecution type: object required: - data - errors properties: errors: type: 'null' data: type: object nullable: false required: - session_id - outputs properties: outputs: type: array items: discriminator: propertyName: type oneOf: - title: StreamOutput description: Outputs that were printed to stdout or stderr type: object required: - type - data properties: type: enum: - stdout - stderr type: string data: type: string - title: ErrorOutput description: >- Errors and exceptions that occurred. If this output type is present, your code did not execute successfully. properties: data: type: string type: enum: - error type: string required: - type - data - properties: data: properties: application/geo+json: type: object additionalProperties: true application/javascript: type: string application/json: type: object additionalProperties: true application/pdf: format: byte type: string application/vnd.vega.v5+json: type: object additionalProperties: true application/vnd.vegalite.v4+json: type: object additionalProperties: true image/gif: format: byte type: string image/jpeg: format: byte type: string image/png: format: byte type: string image/svg+xml: type: string text/html: type: string text/latex: type: string text/markdown: type: string text/plain: type: string type: object type: enum: - display_data - execute_result type: string required: - type - data title: DisplayorExecuteOutput title: InterpreterOutput session_id: type: string description: >- Identifier of the current session. Used to make follow-up calls. example: ses_abcDEF123 nullable: false status: type: string enum: - success description: Status of the execution. Currently only supports success. - title: FailedExecution type: object required: - data - errors properties: data: type: 'null' errors: type: array items: oneOf: - type: string - type: object additionalProperties: true title: Error securitySchemes: bearerAuth: type: http scheme: bearer x-bearer-format: bearer x-default: default ````