Remote MCP server

OpenAI’s Responses API now includes built-in tools for image generation, Code Interpreter, and remote Model Context Protocol (MCP) server support

[{"selector":"#anim-8e549a7a-836e-4a99-9ba4-483e037b3e9e [data-leaf-element=\"true\"]","keyframes":{"transform":["translate3d(12.514648210040342%, 0, 0)","translate3d(0%, 0, 0)"]},"delay":0,"duration":2000,"easing":"cubic-bezier(.3,0,.55,1)","fill":"both"}] [{"selector":"#anim-b8db62f0-df62-48e8-9b5e-db494d5579fc","keyframes":{"opacity":[0,1]},"delay":0,"duration":2000,"easing":"cubic-bezier(.3,0,.55,1)","fill":"both"}] [{"selector":"#anim-e080d770-153b-4c70-afe4-224c1405c147","keyframes":{"transform":["scale(0.3333333333333333)","scale(1)"]},"delay":0,"duration":2000,"easing":"cubic-bezier(.3,0,.55,1)","fill":"forwards"}]

The API supports o-series reasoning models (o1, o3, o3-mini, o4-mini), GPT-4o, and GPT-4.1, enabling advanced agentic applications

[{"selector":"#anim-46de8655-0492-412e-8efb-7069a8def820 [data-leaf-element=\"true\"]","keyframes":{"transform":["translate3d(12.514648210040342%, 0, 0)","translate3d(0%, 0, 0)"]},"delay":0,"duration":2000,"easing":"cubic-bezier(.3,0,.55,1)","fill":"both"}] [{"selector":"#anim-7b3950a7-561f-4b03-878d-5294d5ad21a1","keyframes":{"transform":["translate3d(115.2381%, 0px, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-cf6e92e6-6187-41a4-9a87-b2b0dfc93dd2","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-9684abf6-cc0e-4f70-9938-a51e289d56d1","keyframes":{"transform":["scale(0.15)","scale(1)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"forwards"}]

The new image generation tool (gpt-image-1) supports multi-turn edits and real-time streaming previews, available via the o3 model

[{"selector":"#anim-3e22e81c-685d-458c-8152-5390380cc9eb [data-leaf-element=\"true\"]","keyframes":{"transform":["translate(0%, 0%) scale(1.5)","translate(0%, 0%) scale(1)"]},"delay":0,"duration":2000,"easing":"cubic-bezier(.3,0,.55,1)","fill":"forwards"}] [{"selector":"#anim-f69459a7-9789-47e2-8678-208511413b9a","keyframes":{"transform":["translate3d(-115.2381%, 0px, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-eff19c4b-0f3e-4fd9-ba7a-d2060dfcff04","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-f45a7238-654d-4894-806d-25810b903e53","keyframes":{"transform":["scale(0.15)","scale(1)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"forwards"}]

Code Interpreter is now integrated, enabling data analysis, complex math, coding, and “thinking with images” within the API workflow

[{"selector":"#anim-902743f7-aae5-4afb-bd1f-abb3a34a8bb9 [data-leaf-element=\"true\"]","keyframes":{"transform":["translate3d(12.514648210040342%, 0, 0)","translate3d(0%, 0, 0)"]},"delay":0,"duration":2000,"easing":"cubic-bezier(.3,0,.55,1)","fill":"both"}] [{"selector":"#anim-b7aa9b17-07f0-4d2f-aa00-2dbada3fee88","keyframes":{"transform":["translate3d(115.2381%, 0px, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-858bc7b9-4573-4cc3-a52c-ded026e18ab3","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-25d0a1a6-9157-4d1f-afd3-a2d7b2636a7f","keyframes":{"transform":["scale(0.15)","scale(1)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"forwards"}]

File search improvements include attribute filtering with arrays and support for multiple vector stores, enhancing document retrieval

[{"selector":"#anim-f45489ab-7585-4404-a338-5fea24602183 [data-leaf-element=\"true\"]","keyframes":{"transform":["translate(0%, 0%) scale(1.5)","translate(0%, 0%) scale(1)"]},"delay":0,"duration":2000,"easing":"cubic-bezier(.3,0,.55,1)","fill":"forwards"}] [{"selector":"#anim-f7d790f2-cd2a-4b5d-98f9-495d52079473","keyframes":{"transform":["translate3d(115.2381%, 0px, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-be2d1c69-7af3-485e-9adf-a9d2c835b1d1","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-3b56e378-b83a-4799-adad-1e962f8ed030","keyframes":{"transform":["scale(0.15)","scale(1)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"forwards"}]

o3 and o4-mini models can call tools/functions directly within their reasoning chain, producing richer and more contextually relevant responses

[{"selector":"#anim-49f0a5d7-7ece-4ab8-b506-bbea41f0d3fc [data-leaf-element=\"true\"]","keyframes":{"transform":["translate3d(12.514648210040342%, 0, 0)","translate3d(0%, 0, 0)"]},"delay":0,"duration":2000,"easing":"cubic-bezier(.3,0,.55,1)","fill":"both"}] [{"selector":"#anim-05d9b99a-5f71-40c5-9579-7c443ed57bd0","keyframes":{"transform":["translate3d(115.2381%, 0px, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-57a76905-95f1-4910-9d9d-af73adafe747","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-2af995f0-c021-4759-8cde-0b9189faf425","keyframes":{"transform":["scale(0.15)","scale(1)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"forwards"}]

Reasoning tokens can be preserved across requests and tool calls, reducing latency and developer costs while improving model intelligence

[{"selector":"#anim-a68a1e2f-89cb-4c44-add5-d0814ab652d2 [data-leaf-element=\"true\"]","keyframes":{"transform":["translate3d(12.514648210040342%, 0, 0)","translate3d(0%, 0, 0)"]},"delay":0,"duration":2000,"easing":"cubic-bezier(.3,0,.55,1)","fill":"both"}] [{"selector":"#anim-5c7fc8fe-02b0-4029-80c8-077ddb0702e4","keyframes":{"transform":["translate3d(115.2381%, 0px, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-b304dcf1-411a-4110-ab74-f5e4d118e4af","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-7b885d2b-2246-4987-87df-87e459627ebd","keyframes":{"transform":["scale(0.15)","scale(1)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"forwards"}]

For more details visit Govindhtech.com

[{"selector":"#anim-3cce3b70-cdfe-4749-8613-84d785d683b0 [data-leaf-element=\"true\"]","keyframes":{"transform":["translate(0%, 0%) scale(1.5)","translate(0%, 0%) scale(1)"]},"delay":0,"duration":2000,"easing":"cubic-bezier(.3,0,.55,1)","fill":"forwards"}]