Tech Meridian← ALL MODELS
PROMY MERIDIAN

OPENAI · MODEL RELEASE TRACKER

GPT-6.1 Terra

GPT-6.1 Terra is a mid‑range OpenAI model introduced on 2026-10-26 aimed at bringing fast, capable computer use to everyday work. OpenAI positions it for office tasks and automated software control, combining fast visual reasoning, precise computer control, and a more natural writing style while offering substantially lower API token prices than its predecessor.

CURRENT SNAPSHOT4/5 DIMENSIONS WITH DATA

The dimensions that change the decision.

PRICING
API token pricingOpenAI lists input tokens at $1 per million and output tokens at $5 per million; cached input tokens cost $0.05 per million. OpenAI says these rates are half those of GPT‑5.6 Terra.
CONTEXT WINDOW

Not established from the available sources.

MODALITIES
ModalitiesCombines visual reasoning with text-based outputs and supports precise computer control (multimodal visual + text and agentic control features as described).
BENCHMARKS
Intelligence Index and cost-per-task performanceAt maximum reasoning effort, gains 6 points on the Intelligence Index over GPT-5.6 Terra while costing about $0.40 per task vs $1.40 for GPT-5.6 Terra; at lower effort levels the model gains 7–13 points.
Coding Agent Index v2.0 scoreTerra scores 48 on the Coding Agent Index v2.0 (six points below GPT‑6.1 Sol's 54 and 17 points above GPT‑6 Luna's 31).
Agents' Last Exam performance and token-efficiency comparisonsOpenAI reports Terra matches GPT‑6 Astra on the Agents' Last Exam benchmark at one‑tenth the token price. It is also reported to be more than three times as token‑efficient as Claude Opus 5.5 at maximum effort (at a quarter of the token price), translating to roughly nine times lower cost per task, and to finish more than three times faster than GPT‑6 Astra and GPT‑6.1 Sol.
Reported Intelligence Index improvement and cost per taskOpenAI reports that at maximum reasoning effort Terra gains 6 points on its Intelligence Index over GPT‑5.6 Terra while costing less than a third per task ($0.40 vs $1.40); at lower effort levels reported gains are 7–13 points.
Hallucination/guessing rate (AA‑Omniscience)On the private AA‑Omniscience benchmark Terra resorts to guessing 38% of the time when it lacks an answer, down from 88% for GPT‑5.6 Terra; accuracy also rose by two points (per The Decoder's reporting of OpenAI data).
AVAILABILITY
AvailabilityAvailable now for Plus, Pro, Enterprise, and Edu users in ChatGPT Work and Codex, and via the API.

VERIFIABLE FACTS

Every value stays attached to a source and date.

BENCHMARKS · Agents' Last Exam performance and token-efficiency comparisonsDEVELOPER CLAIM

OpenAI reports Terra matches GPT‑6 Astra on the Agents' Last Exam benchmark at one‑tenth the token price. It is also reported to be more than three times as token‑efficient as Claude Opus 5.5 at maximum effort (at a quarter of the token price), translating to roughly nine times lower cost per task, and to finish more than three times faster than GPT‑6 Astra and GPT‑6.1 Sol.

CAPABILITIES · Target use cases and capabilitiesDEVELOPER CLAIM

Built for office tasks and automated software control; OpenAI describes it as combining fast visual reasoning, precise computer control, and a more natural writing style for tasks such as drafting proposals, interpreting technical drawings, and filling out forms.

WHAT CHANGED

Stored passport versions, without reconstructed history.

Passport updated12 facts
BENCHMARKS · Agents' Last Exam performance and token-efficiency comparisonsOn the Agents' Last Exam benchmark, matches GPT-6 Astra's performance at one‑tenth the token price; OpenAI reports Terra finishes more than three times faster than GPT-6 Astra and GPT-6.1 Sol and is substantially more token‑efficient than Claude Opus 5.5 at maximum effort.→OpenAI reports Terra matches GPT‑6 Astra on the Agents' Last Exam benchmark at one‑tenth the token price. It is also reported to be more than three times as token‑efficient as Claude Opus 5.5 at maximum effort (at a quarter of the token price), translating to roughly nine times lower cost per task, and to finish more than three times faster than GPT‑6 Astra and GPT‑6.1 Sol.
PRICING · API token pricing (per million tokens)− Input tokens: $1 per million; output tokens: $5 per million. Cached inputs: $0.05 per million. OpenAI states these prices are half those of GPT-5.6 Terra.
PRICING · API token pricing+ OpenAI lists input tokens at $1 per million and output tokens at $5 per million; cached input tokens cost $0.05 per million. OpenAI says these rates are half those of GPT‑5.6 Terra.
AVAILABILITY · Availability+ Available now for Plus, Pro, Enterprise, and Edu users in ChatGPT Work and Codex, and via the API.
BENCHMARKS · Coding Agent Index v2.0 scoreScores 48 on the Coding Agent Index v2.0 (six points below GPT-6.1 Sol at 54 and 17 points above GPT-6 Luna at 31).→Terra scores 48 on the Coding Agent Index v2.0 (six points below GPT‑6.1 Sol's 54 and 17 points above GPT‑6 Luna's 31).
BENCHMARKS · Hallucination/guessing rate (AA‑Omniscience)+ On the private AA‑Omniscience benchmark Terra resorts to guessing 38% of the time when it lacks an answer, down from 88% for GPT‑5.6 Terra; accuracy also rose by two points (per The Decoder's reporting of OpenAI data).
BENCHMARKS · Reported Intelligence Index improvement and cost per task+ OpenAI reports that at maximum reasoning effort Terra gains 6 points on its Intelligence Index over GPT‑5.6 Terra while costing less than a third per task ($0.40 vs $1.40); at lower effort levels reported gains are 7–13 points.
CAPABILITIES · Target use cases and capabilities+ Built for office tasks and automated software control; OpenAI describes it as combining fast visual reasoning, precise computer control, and a more natural writing style for tasks such as drafting proposals, interpreting technical drawings, and filling out forms.
AVAILABILITY · Availability− Available now for Plus, Pro, Enterprise, and Edu users in ChatGPT Work and Codex, and available via the API.
RELEASE · Initial release/announcement dateAnnounced/released October 26, 2026→Introduced/announced by OpenAI on 2026-10-26.
Passport updated9 facts
BENCHMARKS · Agents' Last Exam and token-efficiency comparisons+ On the Agents' Last Exam benchmark, matches GPT-6 Astra's performance at one‑tenth the token price; OpenAI reports Terra finishes more than three times faster than GPT-6 Astra and GPT-6.1 Sol and is substantially more token‑efficient than Claude Opus 5.5 at maximum effort.
PRICING · API token pricing (per million tokens)+ Input tokens: $1 per million; output tokens: $5 per million. Cached inputs: $0.05 per million. OpenAI states these prices are half those of GPT-5.6 Terra.
AVAILABILITY · Availability (as stated)− Announcement states the model is intended to bring its capabilities "to everyone."
BENCHMARKS · Coding Agent Index v2.0 score+ Scores 48 on the Coding Agent Index v2.0 (six points below GPT-6.1 Sol at 54 and 17 points above GPT-6 Luna at 31).
BENCHMARKS · Intelligence Index and cost-per-task performance+ At maximum reasoning effort, gains 6 points on the Intelligence Index over GPT-5.6 Terra while costing about $0.40 per task vs $1.40 for GPT-5.6 Terra; at lower effort levels the model gains 7–13 points.
MODALITIES · Modalities+ Combines visual reasoning with text-based outputs and supports precise computer control (multimodal visual + text and agentic control features as described).
AVAILABILITY · Availability+ Available now for Plus, Pro, Enterprise, and Edu users in ChatGPT Work and Codex, and available via the API.
RELEASE · Release dateIntroduced on 2026-10-26 (announcement published 2026-10-26).→Announced/released October 26, 2026
CAPABILITIES · Targeted capabilities and use cases+ Built for everyday office tasks and automated software control; combines fast visual reasoning, precise computer control, and a more natural writing style for tasks like drafting proposals, interpreting technical drawings, and filling out forms.
Passport created3 facts