GPT-6.1 Terra: 50% Cheaper, 3x Faster Computer Use
OpenAI released GPT-6.1 Terra on October 26, 2026, positioning it as the efficiency choice for computer use, visual reasoning, and high-volume professional work. Priced at $1 per million input tokens and $5 per million output tokens—50% lower than GPT-6.1 Sol—Terra scores 15 percentage points higher than its predecessor GPT-5.6 Terra on agent benchmarks while completing tasks over three times faster than GPT-6 Astra and Sol.
- 50% lower pricing — Terra costs $1/M input and $5/M output, half the price of GPT-5.6 Terra and GPT-6.1 Sol, making it practical for high-volume workflows.
- 3x faster on computer use — Terra completes agent tasks over three times faster than GPT-6 Astra and GPT-6.1 Sol on OpenAI's benchmarks.
- Major factuality gains — Error rate dropped from 53.6% (GPT-5.6 Terra) to 16.7% (GPT-6.1 Terra) at low reasoning effort—a 37-point improvement.
What Terra is
Related guide: GPT-6.1 Terra: OpenAI's $1/$5 Model Matches Astra at One-Tenth the Price
Terra is OpenAI's medium-tier model in the GPT-6 lineup, between Luna (cheapest) and Sol (general-purpose). GPT-6.1 Terra represents a significant upgrade from GPT-5.6 Terra, skipped in the initial GPT-6 rollout last month.
OpenAI calls it "a fast, efficient model combining detailed visual reasoning and precise computer use with clearer, more natural writing." The emphasis: professional work tasks—drafting proposals, interpreting technical drawings, filling forms, editing videos.
Pricing and speed advantage
Related guide: Meta Muse Wonder: Meta's First Frontier AI Model Arrives at $4/$20
| Model | Input /1M | Output /1M | Cached /1M |
|---|---|---|---|
| GPT-6.1 Terra | $1.00 | $5.00 | $0.05 |
| GPT-5.6 Terra | $2.00 | $10.00 | $0.10 |
| GPT-6.1 Sol | $2.00 | $10.00 | $0.10 |
| GPT-6 Astra | $8.00 | $40.00 | $0.40 |
That 50% cut matters for high-volume automation—workflows processing thousands of documents, images, or forms daily. Example: reading invoices and entering details into accounting software. At this price, per-task cost becomes acceptable for repetitive work.
Speed compounds savings. On Agents' Last Exam, Terra completes work over 3x faster than GPT-6 Astra and Sol. At max reasoning, Terra is ~9x cheaper per task than Claude Opus 5.5 once speed and tokens combine.
Computer use and visual benchmarks
Related guide: OpenAI Dots: always-on AI agents with their own cloud PC — what actually ships
Agents' Last Exam v1
Tasks like turning architectural drawings into 3D models. At medium reasoning, GPT-6.1 Terra scores 15 points higher than GPT-5.6 Terra at one-third the cost. Matches GPT-6 Astra at one-tenth the price per token.
MMMU-Pro
Terra outperforms GPT-5.6 Terra at every reasoning setting, nearly matching Astra and Sol at lower cost.
OSWorld 2.1
Tests desktop agent tasks. At medium reasoning, Terra surpasses GPT-5.6 Terra's best score at a fraction of cost per task.
Factuality and coding
Related guide: OpenAI agents disabled their own safety monitoring. One then lied about it.
On OpenAI's hallucination benchmark:
- GPT-5.6 Terra: 53.6% error rate (low reasoning)
- GPT-6.1 Terra: 16.7% error rate—drop of 37 points
- At max effort: 4.3%, nearly matching Astra and Sol
On Terminal-Bench 4.0, Terra solves 50.3% of tasks—more than double GPT-5.6 Terra's 21.5%. For most complex coding, Sol and Astra remain better.
Safety and availability
Related guide: Claude Duet: Anthropic's proactive agent that messages you first
Terra meets Critical threshold in cybersecurity under OpenAI's Preparedness Framework. The system card reports improvements in scope adherence, limitation disclosure, and prompt injection resistance vs GPT-5.6 Terra.
Available October 26 to Plus, Pro, Enterprise, Edu in ChatGPT Work and Codex. API: gpt-6.1-terra.
Where Terra fits
Related guide: The UK can test frontier AI models again: what the 17-day freeze and the U-turn actually changed
- Luna—lightweight, cheapest
- Terra—visual reasoning and computer use, mid-tier
- Sol—general reasoning and coding
- Astra—frontier flagship
Terra's sweet spot: agent workflows needing vision and desktop control without Astra-level reasoning. The 3x speed means processing three invoices, drawings, or forms while Sol or Astra handles one—at half the token cost.
For coding/reasoning: Sol remains better. Terra's 50.3% Terminal-Bench is strong, but OpenAI says Sol and Astra preferred for complex coding.
For the frontier: Terra redefines cost-intelligence for computer use. No other model at $1/$5 delivers comparable agent performance with native vision and desktop control.
Bottom line: GPT-6.1 Terra is OpenAI's efficiency play—50% cheaper than Sol, 3x faster than Astra, built for high-volume professional workflows where an agent sees, interprets, and acts. If your automation needs vision and desktop control at scale, Terra just moved the math.
Frequently asked questions
50% lower pricing
Terra costs $1/M input and $5/M output, half the price of GPT-5.6 Terra and GPT-6.1 Sol, making it practical for high-volume workflows.
3x faster on computer use
Terra completes agent tasks over three times faster than GPT-6 Astra and GPT-6.1 Sol on OpenAI's benchmarks.
Major factuality gains
Error rate dropped from 53.6% (GPT-5.6 Terra) to 16.7% (GPT-6.1 Terra) at low reasoning effort—a 37-point improvement.
- OpenAI—Introducing GPT-6.1 Terra (October 26, 2026)
- OpenAI Deployment Safety Hub—GPT-6.1 Terra System Card (October 26, 2026)
- BleepingComputer—OpenAI launches GPT-6.1 Terra with 3x faster AI agents, 50% lower pricing (October 26, 2026)
- 9to5Mac—OpenAI upgrades ChatGPT and Codex with GPT-6.1 Terra (October 26, 2026)
Prices and benchmark figures are standard published rates as of November 2, 2026 and can change without notice. Some outbound links may be affiliate links — see our disclosure. Found an error? Tell us and we'll correct it in public.