supported
Only provider-documented reasoning controls are shown; missing controls remain unpublished rather than inferred.
MODEL REFERENCE / OPENAI
GPT-Realtime-2.1
verified model reference.
OpenAI full-size realtime reasoning and tool-use model for responsive, interruptible production voice agents.
128,000context tokens
32,000max output
2026-10-12last verified
PRIMARY-SOURCE VERIFIED
GPT-Realtime-2.1 at a glance.
OpenAI documents a 128,000-token context window. Input modalities: text + image + audio. Output: text + audio. Pricing status: Official paid API.
MODEL IDgpt-realtime-2.1
CONTEXT WINDOW128,000tokens
MAX OUTPUT32,000
KNOWLEDGE CUTOFF2024-09-30
STANDARD INPUT$4.00per 1 million TEXT tokens (audio/image separate)
STANDARD OUTPUT$24.00per 1 million TEXT tokens (audio/image separate)
CAPABILITY CONTRACT
Published model behavior.
text + image + audio
Input modalities come from the cited official model documentation.
text + audio
Output modality and output-limit claims remain separate so an undocumented token cap is not guessed.
PRICING STATUS
Official paid API.
$4.00 input · $0.40 cached · $24.00 output per 1 million TEXT tokens (audio/image separate).
SXF preserves the provider-published native billing basis. Only calculator-eligible generative token models enter token-cost arithmetic; specialist units and input-only embeddings remain visible without forced conversion.
CAVEATS
What the source record does not assume.
01
Displayed Standard token rates are TEXT ONLY: $4.00 input, $0.40 cached, and $24.00 output per 1M text tokens.
02
AUDIO Standard rates differ: $32.00 input, $0.40 cached input and $64.00 output per 1M audio tokens. IMAGE input is $5.00, cached $0.50 per 1M image tokens.
03
Mixed-modality realtime bills each modality separately; the generic text cost calculator must not be used to estimate full voice-session bills.
04
The model supports configurable reasoning effort, which can affect token usage and response latency.
EXPLORE
Keep the model in context.
OFFICIAL SOURCES
Verification links.
WHAT CHANGED
GPT-Realtime-2.1: what changed.
No post-baseline factual changes recorded for GPT-Realtime-2.1 since 2026-10-12.
VERIFIED BASELINE2026-10-12Current facts remain aligned with the SXF ledger.
VERIFIED CHANGE HISTORY
GPT-Realtime-2.1 verified history.
GPT-Realtime-2.1
Baseline verified: 128,000 context · 32,000 max output · $4.00 input / $0.40 cached / $24.00 output per 1 million TEXT tokens (audio/image separate)
Official evidence ↗Append-only SXF ledger · 1 verified event · chain head b12eeb63a07e…