OpenAI logo

GPT-6.1 Terra

Multimodal
OpenAI

GPT-6.1 Terra is OpenAI's fast, efficient GPT-6.1 model for visual reasoning, precise computer use, and clear, natural writing, priced 50% below GPT-5.6 Terra and GPT-6.1 Sol. The API model id is gpt-6.1-terra. It accepts text and image input with text output, supports none, low, medium (API default), high, xhigh, and max reasoning effort, and provides a 1,050,000-token context window with up to 128,000 output tokens. Use the Responses API for tool calling; Chat Completions is supported without tool calling. Responses API tools include web and file search, image generation, code interpreter, hosted shell, apply patch, skills, computer use, MCP, and tool search; it also supports streaming, function calling, and structured outputs. Standard per-million-token pricing is $1 input, $0.05 cached input, $1.25 cache writes (30-minute retention rate; 12-hour retention is $2.50), and $5 output. Prompts above 272K input tokens cost 2x input/cache and 1.5x output across the full request. Batch and Flex cost 50% of Standard; Fast costs 2x. Regional processing adds 10% where available. Launch benchmarks are OpenAI-reported; effort and evaluation variants are recorded with each score.

Key Specifications

Parameters
-
Context
1.1M
Release Date
October 26, 2026
Average Score
53.3%

Timeline

Key dates in the model's history
Announcement
October 26, 2026
Last Update
October 31, 2026
Today
November 20, 2026

Technical Specifications

Parameters
-
Training Tokens
-
Knowledge Cutoff
June 1, 2026
Family
-
Capabilities
MultimodalZeroEval

Pricing & Availability

Input (per 1M tokens)
$1.00
Output (per 1M tokens)
$5.00
Max Input Tokens
1.1M
Max Output Tokens
128.0K
Supported Features
Function CallingStructured OutputCode ExecutionWeb SearchBatch InferenceFine-tuning

Benchmark Results

Model performance metrics across various tests and benchmarks

Other Tests

Specialized benchmarks
Agents' Last Exam
Agents' Last Exam v1. Self-reported by OpenAI at max reasoning effort (0.600), GPT-6.1 Terra launch interactive chart. Other efforts: low 0.559, medium 0.576, high 0.572, xhigh 0.585. Cost per task at max effort: $1.01. Evaluations used OpenAI research environments or API and may differ from production ChatGPT. • Self-reported
60.0%
HealthBench
HealthBench, length-adjusted score (unadjusted 0.511, mean response length 1,331 chars). GPT-6.1 Terra system card addendum Table 5. Evaluations used OpenAI research environments or API and may differ from production ChatGPT. • Self-reported
55.1%
HealthBench Consensus
HealthBench Consensus, length-adjusted score (unadjusted 0.948, mean response length 1,312 chars). GPT-6.1 Terra system card addendum Table 5. Evaluations used OpenAI research environments or API and may differ from production ChatGPT. • Self-reported
95.1%
HealthBench Hard
HealthBench Hard, length-adjusted score (unadjusted 0.260, mean response length 1,306 chars). GPT-6.1 Terra system card addendum Table 5. Evaluations used OpenAI research environments or API and may differ from production ChatGPT. • Self-reported
31.4%
HealthBench Professional
HealthBench Professional, length-adjusted score (unadjusted 0.642, mean response length 2,531 chars). GPT-6.1 Terra system card addendum Table 5. Evaluations used OpenAI research environments or API and may differ from production ChatGPT. • Self-reported
62.7%
MentalHealthBench
MentalHealthBench overall at max reasoning effort (0.555 ± 0.010). Acuity splits at max effort: non-acute 0.565, high acuity 0.565, emergent 0.529. GPT-6.1 Terra system card addendum Table 7. Evaluations used OpenAI research environments or API and may differ from production ChatGPT. • Self-reported
55.5%
Terminal-Bench 4.0
Terminal-Bench 4.0. OpenAI GPT-6.1 Terra launch post: at max reasoning effort GPT-6.1 Terra solves 50.3% of tasks (GPT-5.6 Terra published 21.5%). Evaluations used OpenAI research environments or API and may differ from production ChatGPT. • Self-reported
50.3%
Terminal-Bench-Science 0.1
Terminal-Bench-Science 0.1. Self-reported by OpenAI at max reasoning effort (0.583), GPT-6.1 Terra launch interactive chart. Other efforts: low 0.243, medium 0.343, high 0.397, xhigh 0.434. Evaluations used OpenAI research environments or API and may differ from production ChatGPT. • Self-reported
58.3%
LifeSci Diamond (2026-11)
Strict pass rate on 55 regulatory-affairs tasks. Plain API via Vercel AI Gateway with a web_search tool (Exa), provider default settings, one answer per task, LLM-jury grading (Arca, 2026-11). • Verified
10.9%

License & Metadata

License
proprietary
Announcement Date
October 26, 2026
Last Updated
October 31, 2026

Similar Models

All Models

Recommendations are based on similarity of characteristics: developer organization, multimodality, parameter size, and benchmark performance. Choose a model to compare or go to the full catalog to browse all available AI models.