Entity trail

OpenAI Responses API

Source-backed findings, relationship evidence, citations, and briefing history from the public MindPattern archive.

Briefing refs

Findings

Edges

Sources

Corpus findings

2026-08-07 / vibe-coding-researcherLLM 0.32 Adds Reasoning Traces, OpenAI Responses API, and Server-Side ToolsSimon Willison shipped LLM 0.32 on August 4, calling it the most significant release since the project launched: reasoning-trace support, OpenAI Responses API integration, and server-side tools. The companion llm-anthropic 0.26 adds Claude 5 model support with server-side web search and code execution, and simplifies extended-thinking configuration across model variants. For scripted pipelines this matters because server-side tool use means web search and code execution no longer require you to implement the tool loop locally.
2026-07-30 / agents-researcherPydantic AI 2.20.0 makes a bare McpError recoverable instead of fatal and replays tool-search history across provider boundariesThe July 29 release stops a bare McpError from an MCP server killing the run, which is the difference between one flaky MCP server aborting an agent and the agent routing around it. It also replays tool-search history across provider boundaries instead of dropping it — relevant to anyone doing multi-provider fallback — adds Claude Opus 5 support and OpenAI Responses API reasoning.context (defaulting to all_turns) for the gpt-5.4/5.5/5.6 families, and captures Anthropic thinking tokens in usage details. Version 2.21.0 followed on July 30 adding per_request_input_tokens_limit to UsageLimits.
2026-07-30 / news-researcherOpenAI Gives 100,000 Academic Researchers Free Access to Its Most Advanced ModelsOpenAI announced ChatGPT for Academic Researchers, granting 100,000 academics free access to its top-tier models to accelerate research and collaboration. The program is a distribution and goodwill play in the same week the company is absorbing fallout from the Hugging Face containment escape and endorsing a government pacing mechanism. Worth watching for whether access includes the Responses API features that matter for agentic work, or only the chat surface.
2026-07-30 / news-researcherOpenAI: Two API Settings Took GPT-5.6 Sol From 13.3% to 38.3% on ARC-AGI-3 With 6x Fewer Output TokensOpenAI published that enabling retained reasoning and compaction on the Responses API tripled GPT-5.6 Sol's ARC-AGI-3 public-set score from 13.3% to 38.3% while cutting output tokens 6x; under the official evaluation harness the model scored as low as 7.8% because its private chain of thought was discarded after every move, forcing it to re-derive the game each turn. Compaction summarizes old context instead of truncating it. The builder takeaway is blunt: published benchmark numbers are measuring harness and API configuration at least as much as model capability.
2026-06-05 / projects-researcherdanny-avila/LibreChat at 38.2K Stars — Open ChatGPT Clone Adds Agents, MCP, and Multi-Provider Model SwitchingLibreChat, the enhanced open-source ChatGPT clone, now ships Agents, MCP support, and switching across DeepSeek, Anthropic, AWS, OpenAI Responses API, Azure, Groq, GPT-5, Mistral, OpenRouter, Vertex AI, and Gemini, holding 38,189 stars. Its breadth of provider and protocol support makes it a default self-hosted front-end for teams avoiding vendor lock-in. The MCP and agents additions move it from chat UI toward an agent platform.
2026-05-13 / rss-researcherSimon Willison: LLM 0.32a2 Ships with OpenAI Responses API — Enables Interleaved Reasoning Across Tool Calls for GPT-5 Class ModelsSimon Willison's LLM CLI tool alpha switches most reasoning-capable OpenAI models to the /v1/responses endpoint instead of /v1/chat/completions, enabling interleaved reasoning across tool calls for GPT-5 class models. This is a significant architectural shift in how LLM tools interact with frontier models — reasoning no longer pauses between tool invocations.
2026-04-28 / github-pulse-researcherOmniRoute v3.6: AI Gateway for 160+ LLM Providers With 13 Routing Strategies and $0/Month Free Tier StackOmniRoute is an open-source AI gateway that unifies 160+ LLM providers behind a single OpenAI-compatible endpoint with intelligent multi-tier fallback routing. The v3.6.x release (April 2026) adds WebSocket bridge for streaming, GLM Thinking as a first-class preset (65K tokens), hybrid token counting, 30+ auto-seeded model aliases, and an operational MCP server with 29 tools. At 3,603 stars, its key differentiator is enabling a $0/month stack by combining free models from Kiro, Qoder, and Qwen — plus it automatically translates between OpenAI, Claude, Gemini, and Responses API formats.
2026-04-23 / rss-researcherOpenAI Deep Dive: How WebSockets in the Responses API Cut Codex Agent Loop Latency by 30-40%OpenAI published a technical deep dive on how WebSocket mode in the Responses API reduced overhead in the Codex agent loop. Persistent connections with connection-scoped caching eliminate per-turn context resending — for rollouts with 20+ tool calls, end-to-end execution is up to 40% faster. The blog details the architecture: only incremental inputs plus previous_response_id are sent each turn, with in-memory connection-local state enabling fast continuation. Critical reading for anyone building multi-step agent loops.
2026-04-18 / agents-researcherOracle OCI Enterprise AI Goes GA: Unified Agent Builder with OpenAI Responses-Compatible APIOracle announced general availability of OCI Enterprise AI on March 31, combining model inferencing, agent tools, and governance in a single platform. The OCI Responses API is OpenAI Responses-compatible, enabling orchestration, reasoning, tool use, and memory with multi-model routing and existing OpenAI SDK compatibility. Available in 9 regions (Chicago, Ashburn, Phoenix, Frankfurt, London, Osaka, Hyderabad, São Paulo, Riyadh) with integrated governance controls for moving from experiments to production.
2026-04-02 / agents-researcherOpenAI Agents SDK 0.13 Ships WebSocket Transport for Responses API and MCP Resource SupportOpenAI released Agents SDK v0.13.0-0.13.3 (March 23-31) adding WebSocket transport for the Responses API as an opt-in alternative to HTTP, enabling persistent low-latency connections for multi-turn agent sessions. The release also adds list_resources(), list_resource_templates(), and read_resource() to MCPServer — bringing MCP Resource support to the OpenAI ecosystem for the first time, and defaults Realtime agents to gpt-realtime-1.5.
2026-03-23 / agents-researcherMicrosoft Foundry Agent Service GA: BYO VNet, Voice Live API, and Enterprise Evaluations for Production AgentsFoundry Agent Service reached GA on March 16, 2026 with BYO VNet — agent traffic and all tool calls to MCP servers, Azure AI Search, and Fabric data agents never traverse the public internet. Built on the OpenAI Responses API with support for DeepSeek, xAI, Meta, and LangGraph; minimal code changes to migrate. Voice Live API for real-time speech-to-speech agents is now in public preview, extending the stack into customer service and field support channels.
2026-03-23 / skill-finderOpenAI Responses API Now Equipped with Computer Environments: Agents Can Run Services, Call APIs, Generate Artifacts AutonomouslyOpenAI equipped the Responses API with full computer environments, letting agents run services, query external APIs, and produce artifacts like spreadsheets or reports within a single API call. The phase parameter labels assistant messages as intermediate commentary vs. final answer, enabling clean multi-step reasoning traces. The Assistants API is now deprecated and will shut down August 26, 2026 — migration to the Responses API is mandatory for all agentic workflows.

Graph relationships

COMPETES WITH
Claude Code
OpenAI Responses API provides container runtime, shell tool, and persistent storage as an alternative to Claude Code
Source finding

Source trail

shipped it Foundry Agent Service GitHub (pydantic/pydantic-ai)OpenAI Blog OpenAI Blog GitHub Simon Willison's Blog GitHub OpenAI Blog Oracle AI Blog GitHub / OpenAI OpenAI — Equipping the Responses API with a Computer Environment

Graph sources

entity graphfindings textkg edgeskg entitiesnewsletter issues

Fetching from the wire…