GPT-Realtime-2.1

by OpenAI · gpt family · released 2026-07-06

On output pricing, GPT-Realtime-2.1 is cheaper than 59 of the 613 priced models tracked here, at $24/Mtok. Its 128,000-token context window is larger than 191 of 701 models. Capabilities: vision, tools, reasoning.

Specifications

ProviderOpenAI
Familygpt
Input price$4/Mtok
Output price$24/Mtok
Cache read price$0.4/Mtok
Context window128,000
Max output tokens32,000
Input modalitiestext, audio, image
Output modalitiestext, audio
Capabilitiesvision, tools, reasoning
Knowledge cutoff2024-09-30
Released2026-07-06

Estimate GPT-Realtime-2.1 costs for your workload →

Similar models

GPT-5.6
OpenAI · $30/Mtok out · 1.1M context
GPT-5.6 Sol
OpenAI · $30/Mtok out · 1.1M context
GPT-5.5
OpenAI · $30/Mtok out · 1.1M context
GPT-5.4
OpenAI · $15/Mtok out · 1.1M context

Related reading

OpenAI's GPT-5.6 Sol, Terra, and Luna: What Just Launched (and Why You Can't Use It Yet) OpenAI Codex Sites Just Launched: What Indie Hackers Actually Need to Know Claude Opus 4.8 vs GPT-5.5 for Indie Hackers in 2026: The Matchup Just Flipped
OpenAI documentation → ← All AI models

Data synced daily from models.dev (last synced 1 day ago). Always verify pricing with OpenAI before deploying to production.