News

GPT-6 Astra Lands in ComfyUI: A Million-Token Context Inside the Graph

18 September 2026 · RenderBob team

OpenAI's GPT-6 Astra, released 3 September 2026, is now usable inside ComfyUI through the OpenAI Chat node, with a 1,050,000-token context and adjustable reasoning effort.

OpenAI's GPT-6 Astra, released 3 September 2026 as the company's most capable model, built for complex reasoning, coding, computer use, research and document creation, is now usable directly inside ComfyUI through the existing OpenAI Chat node. Select gpt-6-astra in the model dropdown and it is in the graph, with a 1,050,000-token context window, text and image input, and reasoning effort from low through max.

Local LLM nodes already drive prompt refinement, routing decisions, and multi-step agentic logic inside production graphs, not just chat. Astra's additions matter for that role. A 1.05-million-token window can hold a project's brief, brand guidelines, prior approved prompts, and a long shot list at once, instead of the truncation and re-summarising that smaller-context models force. Adjustable reasoning effort is a cost-and-latency lever: low for a quick prompt touch-up mid-iteration, max for a hard continuity brief, with the studio choosing the trade-off per node.

The setup matches the rest of ComfyUI's partner nodes: confirm you are logged in and on a permitted network, keep ComfyUI updated, and use the Workflow Templates panel if the node does not appear. A missing node at load time is usually an outdated version or an import failure at startup.

A reasoning model this capable, wired into production logic, belongs in the same approved-model registry and reproducibility discipline as an image or video model. It is also an API dependency: per-call cost, data leaving your environment on every call, and a model whose retirement schedule you do not control. This blog's coverage of the Kling v2 retirement is the practical picture of that risk. Use Astra on the reasoning-heavy nodes where the scale helps, and put it through the same cost-and-governance rules as everything else in the graph.