# Question about gpt-5 batch API **URL:** \ **Category:** API\ **Created:** [October 11, 2025, 6:54am UTC](https://community.openai.com/t/question-about-gpt-5-batch-api/1361972 "2025-10-11T06:54:24Z")\ **Posts on this page:** 3\ **Page:** 1
**Author:** ![Hijkstuv](https://sea2.discourse-cdn.com/openai1/user_avatar/community.openai.com/hijkstuv/32/688630_2.png) [@Hijkstuv](https://community.openai.com/u/Hijkstuv)\ **Post date:** [October 11, 2025, 6:54am UTC](https://community.openai.com/t/question-about-gpt-5-batch-api/1361972/1 "2025-10-11T06:54:25Z")
Does anyone know how to apply structured output when using batch API posting on \v1\responses, with gpt-5? ---
**Author:** ![sps](https://sea2.discourse-cdn.com/openai1/user_avatar/community.openai.com/sps/32/7736_2.png) [@sps](https://community.openai.com/u/sps)\ **Post date:** [October 11, 2025, 7:55am UTC](https://community.openai.com/t/question-about-gpt-5-batch-api/1361972/2 "2025-10-11T07:55:38Z")
Welcome to the dev community @Hijkstuv Here’s an example of a one line JSON for structured outputs over Responses API using the Batch API: ```auto {"custom_id": "request001", "method": "POST", "url": "/v1/responses", "body": { "model": "gpt-5", "input": [{ "role": "system", "content": "You are an expert at structured data extraction. You will be given unstructured text from a research paper and should convert it into the given structure." }, { "role": "user", "content": "..." }], "text": { "format": { "type": "json_schema", "name": "research_paper_extraction", "schema": {"type": "object","properties": {"title": { "type": "string" },"authors": { "type": "array", "items": { "type": "string" }},"abstract": { "type": "string" },"keywords": { "type": "array", "items": { "type": "string" }}},"required": ["title", "authors", "abstract", "keywords"],"additionalProperties": false }, "strict": true } }}} ``` This was put together simply by placing the body of the cURL request in the [Structured output examples from docs](https://platform.openai.com/docs/guides/structured-outputs#examples) into the body of a Batch API input request for the Responses API. ---
**Author:** ![\_j](https://sea2.discourse-cdn.com/openai1/user_avatar/community.openai.com/_j/32/766292_2.png) [@\_j](https://community.openai.com/u/_j)\ **Post date:** [October 11, 2025, 10:48am UTC](https://community.openai.com/t/question-about-gpt-5-batch-api/1361972/3 "2025-10-11T10:48:06Z")
Here are some data employing the above line, prepared in Python: ```python import json line = r'''that_line_copied''' request_dict = json.loads(line) pretty_json = json.dumps(request_dict, indent=2) compact_line = json.dumps(request_dict, separators=(',', ':')) print(f"{pretty_json}\n\n{compact_line}") ``` The `request_dict.body` is like you might already use in a normal request that serializes Python objects to a JSON-formatted string request body: `openai_client.responses.create(**request_dict["body"])` `pretty_json` is so we can see what is being communicated on the forum. `compact_line` is a string to add to a JSONL, but without excess bytes. ```json { "custom_id": "request001", "method": "POST", "url": "/v1/responses", "body": { "model": "gpt-5", "input": [ { "role": "system", "content": "You are an expert at structured data extraction. You will be given unstructured text from a research paper and should convert it into the given structure." }, { "role": "user", "content": "..." } ], "text": { "format": { "type": "json_schema", "name": "research_paper_extraction", "schema": { "type": "object", "properties": { "title": { "type": "string" }, "authors": { "type": "array", "items": { "type": "string" } }, "abstract": { "type": "string" }, "keywords": { "type": "array", "items": { "type": "string" } } }, "required": [ "title", "authors", "abstract", "keywords" ], "additionalProperties": false }, "strict": true } } } } ``` ```json {"custom_id":"request001","method":"POST","url":"/v1/responses","body":{"model":"gpt-5","input":[{"role":"system","content":"You are an expert at structured data extraction. You will be given unstructured text from a research paper and should convert it into the given structure."},{"role":"user","content":"..."}],"text":{"format":{"type":"json_schema","name":"research_paper_extraction","schema":{"type":"object","properties":{"title":{"type":"string"},"authors":{"type":"array","items":{"type":"string"}},"abstract":{"type":"string"},"keywords":{"type":"array","items":{"type":"string"}}},"required":["title","authors","abstract","keywords"],"additionalProperties":false},"strict":true}}}} ``` To ensure success, I’m going to directly consume from the `compact_line` meant for a batch: ```python import os, httpx, json try: with httpx.Client(timeout=1000) as client: resp = client.post( "https://api.openai.com/" + json.loads(compact_line).get('url'), headers={ "Authorization": f"Bearer {os.getenv('OPENAI_API_KEY')}", "Content-Type": "application/json", "Accept": "application/json", }, content=json.dumps(json.loads(compact_line).get('body')), ) resp.raise_for_status() except httpx.HTTPStatusError as e: print(f"Request failed: {e}") if e.response is not None: try: # print body error messages from OpenAI print("Error response body:\n", e.response.text) except Exception: pass raise except httpx.RequestError as e: print(f"Request error: {e}") raise ``` `` * * * We see below the `text.format` with `"type": "json_schema"` echoed back at you in the large response object (that comes pretty-formatted), along with an AI forced to use that schema even though there is nothing to “say” about the user’s ellipsis input. `print(resp.text)` ```json { "id": "resp_1234", "object": "response", "created_at": 1760179057, "status": "completed", "background": false, "billing": { "payer": "openai" }, "error": null, "incomplete_details": null, "instructions": null, "max_output_tokens": null, "max_tool_calls": null, "model": "gpt-5-2025-08-07", "output": [ { "id": "rs_1234", "type": "reasoning", "summary": [] }, { "id": "msg_1234", "type": "message", "status": "completed", "content": [ { "type": "output_text", "annotations": [], "logprobs": [], "text": "{\"title\":\"\",\"authors\":[],\"abstract\":\"\",\"keywords\":[]}" } ], "role": "assistant" } ], "parallel_tool_calls": true, "previous_response_id": null, "prompt_cache_key": null, "reasoning": { "effort": "medium", "summary": null }, "safety_identifier": null, "service_tier": "default", "store": true, "temperature": 1.0, "text": { "format": { "type": "json_schema", "description": null, "name": "research_paper_extraction", "schema": { "type": "object", "properties": { "title": { "type": "string" }, "authors": { "type": "array", "items": { "type": "string" } }, "abstract": { "type": "string" }, "keywords": { "type": "array", "items": { "type": "string" } } }, "required": [ "title", "authors", "abstract", "keywords" ], "additionalProperties": false }, "strict": true }, "verbosity": "medium" }, "tool_choice": "auto", "tools": [], "top_logprobs": 0, "top_p": 1.0, "truncation": "disabled", "usage": { "input_tokens": 98, "input_tokens_details": { "cached_tokens": 0 }, "output_tokens": 536, "output_tokens_details": { "reasoning_tokens": 512 }, "total_tokens": 634 }, "user": null, "metadata": {} } ``` Verifying @sps gave us a working line (just with a few extra spaces). So the answer to the question is that, “Yes, user sps knows”.