{"slug": "openais-gpt-6-1-terra-max-lands-at-8-in-agent-arena-with-a-10-6-net-improvement", "title": "OpenAI’s GPT-6.1 Terra (Max) lands at #8 in Agent Arena with a 10.6% net improvement", "summary": "OpenAI's GPT-6.1 Terra, running at its Max setting, placed #8 in Agent Arena with a 10.6% net improvement and a median task cost of $0.62, according to the ranking. OpenAI released GPT-6.1 Terra on October 26, 2026, priced at $1 per million input tokens and $5 per million output tokens, roughly 50% cheaper than GPT-6.1 Sol and GPT-5.6 Terra, with a 1.05 million token context window and availability through ChatGPT Work, Codex, and the API under the model ID gpt-6.1-terra. The model scored 50.3% on Terminal-Bench 4.0 versus 21.5% for GPT-5.6 Terra at maximum effort, a 60% pass rate on the Agents' Last Exam, and around 74% on OSWorld.", "body_md": "# OpenAI’s GPT-6.1 Terra (Max) lands at #8 in Agent Arena with a 10.6% net improvement\n\nFoxTPNL / Wikimedia Commons (CC BY 4.0)\n\nOpenAI's mid-tier model posts a median task cost of $0.62, making a case that cheaper can still be competitive\n\n[OpenAI](https://cryptobriefing.com/markets/openai/)’s GPT-6.1 [Terra](https://cryptobriefing.com/markets/terra-luna-2/), running at its Max setting, has placed **#8 in Agent Arena**. It posted a 10.6% net improvement and a median task cost of $0.62.\n\n## What the numbers show\n\nAgent Arena ranks AI models on agentic work. These are tasks where the model acts on a user’s behalf rather than simply answering questions. Terra (Max) came in eighth overall.\n\nThe 10.6% net improvement is the headline metric from the ranking. Next to it sits the cost figure: a median of $0.62 per task.\n\nBroader evaluations put Terra’s median task costs between $0.40 and $0.62. The Agent Arena result sits at the top of that range. That fits a Max configuration, which trades extra compute for better output.\n\nThe model has posted solid scores elsewhere as well. On Terminal-Bench 4.0 it scored 50.3%, compared with 21.5% for its predecessor, GPT-5.6 Terra, at maximum effort.\n\nTerra (Max) also achieved a 60% pass rate on the Agents’ Last Exam at max effort. It scored around 74% on OSWorld, a benchmark that measures how well a model can operate a computer the way a person would.\n\n### AI, tech, and the markets they move—in one daily briefing.\n\nDaily. Free. Join 34,000+ readers across crypto, finance, and policy.\n\n## Where Terra fits in the GPT-6 family\n\nOpenAI released GPT-6.1 Terra on October 26, 2026. The company designed it for cost-efficient visual reasoning, agentic tasks, and natural writing.\n\nThe target use cases include clicking through screens, filling out forms, processing invoices, and handling documents.\n\nAPI pricing comes in at $1 per million input tokens and $5 per million output tokens. That makes it approximately 50% more affordable than its closest counterparts, GPT-6.1 Sol and GPT-5.6 Terra.\n\nThe model also reportedly completes tasks more than three times faster than its more expensive siblings.\n\nTerra is available through ChatGPT Work, Codex, and the API under the model ID gpt-6.1-terra. It is not accessible through the standard ChatGPT chat interface.\n\nIt also ships with a context window of 1.05 million tokens.\n\nWithin OpenAI’s GPT-6 series, Terra occupies the middle of the menu. Sol and Astra sit above it as the high-performance models. Luna sits below as the lighter option.\n\n**Disclosure:** This article was edited by Diego Almada Lopez. For more information on how we create and review content, see our [Editorial Policy](https://cryptobriefing.com/editorial-policy/).\n\n## More in Technology\n\n[See all](https://cryptobriefing.com/tech/)", "url": "https://wpnews.pro/news/openais-gpt-6-1-terra-max-lands-at-8-in-agent-arena-with-a-10-6-net-improvement", "canonical_source": "https://cryptobriefing.com/openai-gpt-6-1-terra-agent-arena-ranking/", "published_at": "2026-10-29 15:03:43+00:00", "updated_at": "2026-10-29 15:21:30.382792+00:00", "lang": "en", "topics": ["artificial-intelligence", "large-language-models", "ai-agents", "ai-products"], "entities": ["OpenAI", "GPT-6.1 Terra", "Agent Arena", "GPT-6.1 Sol", "GPT-5.6 Terra", "ChatGPT Work", "Codex", "Terminal-Bench 4.0"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/openais-gpt-6-1-terra-max-lands-at-8-in-agent-arena-with-a-10-6-net-improvement", "markdown": "https://wpnews.pro/news/openais-gpt-6-1-terra-max-lands-at-8-in-agent-arena-with-a-10-6-net-improvement.md", "text": "https://wpnews.pro/news/openais-gpt-6-1-terra-max-lands-at-8-in-agent-arena-with-a-10-6-net-improvement.txt", "jsonld": "https://wpnews.pro/news/openais-gpt-6-1-terra-max-lands-at-8-in-agent-arena-with-a-10-6-net-improvement.jsonld"}}