Mid-tier GPT-6 model released 2026-09-29 that OpenAI says nears GPT-6 Astra on coding, computer use and professional work at a fifth of Astra's token price; 1.05M context, low-to-max effort, optional Ultrafast mode.
Most efficient GPT-6 model (September 2026) for focused, high-volume work, priced at $0.10 input and $0.50 output per million tokens, with the full 1.05M context and effort from none to max. OpenAI's suggested replacement for retiring GPT-4o and GPT-4.1 snapshots.
First Sol model of the GPT-6 generation (September 2026), built for complex coding and agentic workflows with a 1.05M context and none-to-max reasoning. Followed a week later by GPT-6.1 Sol, which replaced it in OpenAI's main lineup.
Faster of OpenAI's two September 2026 GPT Image 2.5 models, for everyday generation and editing; OpenAI cites up to 50% lower latency than Images 2.0. Quality settings run from low to max; alias of the 2026-09-08 snapshot.
Dated September 8, 2026 snapshot that the gpt-image-2.5-flare alias currently resolves to, for callers who want Flare's output fixed if the alias moves. Billed at the same token rates as GPT Image 2.
OpenAI's most capable image model as of October 2026 (released September 2026), aimed at precise, detail-heavy generation and editing at the cost of longer generation times; quality from low to max. Alias of the 2026-09-08 snapshot.
Pinned September 8, 2026 snapshot of GPT Image 2.5 Sunburst and the default behind its alias; text and image in, image out, through the generations, edits and Batch endpoints.
OpenAI's flagship as of October 2026, released September 2026 as the top GPT-6 model for demanding reasoning, coding, computer use and research. 1.05M-token context, 128K output, reasoning effort from low to max.
OpenAI's current high-accuracy speech-to-text model (July 2026) for files and committed Realtime turns, accepting context prompts, keyword hints and multiple language hints, at $0.0045 per minute. Named successor to whisper-1 and the GPT-4o transcribe models.
Low-cost tier of OpenAI's GPT-5.6 family (July 2026), roughly the old nano slot, with a 1.05M context and effort settings from none to max. Superseded by GPT-6 Luna in September 2026.
Top model of the GPT-5.6 family, generally available July 2026 and the target of the gpt-5.6 alias; 1.05M context and reasoning effort up to a new max level. Superseded by GPT-6 Astra and GPT-6.1 Sol in September 2026.
Middle tier of GPT-5.6 (July 2026), balancing cost and capability in the slot earlier GPT-5 releases filled with a mini model; 1.05M context, effort up to max. The GPT-6 generation has no Terra tier.
July 2026 update to gpt-realtime-2 with better alphanumeric recognition, silence and noise handling, and interruption behavior; configurable reasoning effort, 128K context. OpenAI's named replacement for gpt-realtime and gpt-4o-realtime.
Distilled, lower-cost version of gpt-realtime-2.1 released July 2026, with configurable reasoning and a 128K context at $10/$20 per 1M audio tokens in/out. OpenAI's named replacement for gpt-realtime-mini, tts-1 and gpt-4o-mini-tts.
Realtime API speech-to-speech model released May 2026, adding configurable reasoning effort and tool use for voice agents, with a 128K context and 32K max output tokens. Updated by gpt-realtime-2.1 in July 2026.
Streaming speech-to-text model released May 2026 for live captions and notes, emitting low-latency transcript deltas in Realtime transcription sessions at $0.017 per minute. Not available on the file transcription endpoint; GPT-Live-Transcribe followed in July 2026.
OpenAI's April 2026 model for agentic coding, computer use and knowledge work, with a 1.05M context, 128K output and none-to-xhigh reasoning. It led OpenAI's lineup until GPT-5.6 in July and GPT-6 Astra in September 2026.
Pro version of GPT-5.5 (April 2026) that spends extra compute for more precise answers; medium, high (default) and xhigh effort, Responses and Batch APIs only. OpenAI's latest standalone Pro model; for GPT-5.6 Sol it offers a pro reasoning mode instead.
OpenAI's April 2026 image generation and editing model, with flexible output sizes and high-fidelity image inputs; alias of the gpt-image-2-2026-04-21 snapshot. Superseded in September 2026 by GPT Image 2.5 Sunburst and Flare.
Pinned April 21, 2026 snapshot behind the gpt-image-2 alias, launched with ChatGPT Images 2.0. Supports inpainting through the edits endpoint and, since August 2026, a transparent-background preview.
