On October 26, OpenAI introduced GPT-6.1 Terra — a model that halves the cost of working with tokens compared to the previous version. However, independent tests show that two Anthropic models still hold the lead in the intelligence ranking.
Terra now costs $1 per million input tokens and $5 per million output tokens. For GPT-5.6 Terra, these figures were $2 and $12 respectively. The cost of solving a single task dropped to $0.40 versus $1.40 for its predecessor — a decline of roughly 70%.
Terra's Position in the Independent Ranking
In the Artificial Analysis Intelligence Index, GPT-6.1 Terra scored 48 points at maximum effort versus 42 points for GPT-5.6 Terra. At the same time, the model now answers "at random" on 38% of questions — previously this figure reached 88%, indicating notable progress in reducing hallucinations.
Nevertheless, Anthropic's models retain the top spots in the ranking. Claude Opus 5.5 leads with 58 points, and Claude Sonnet 5.5 scored 56. Gemini 4 Argon from Google and GPT-6 Astra from OpenAI each received 53 points.
In its announcement, OpenAI emphasizes economics above all: Terra is nine times cheaper than Opus 5.5 when tested on the Agents' Last Exam benchmark. For the most complex programming tasks, the company still recommends GPT-6.1 Sol or GPT-6 Astra.
Anthropic Goes Public
Anthropic filed an S-1 form for an IPO on October 26, planning to list on Nasdaq under the ticker ANT. According to the filing, the company's revenue for the first half of 2026 will be $16.29 billion.
OpenAI is in no hurry to enter the public market: CEO Sam Altman ruled out a stock offering this year, and the company is raising private investment. The target valuation for the new round was lowered to $1.25 trillion.
The reduction in Terra's prices is also reflected in OpenAI's revenue. To maintain its previous revenue in this segment, the company needs to sell roughly 3.5 times as many tasks. Terra is specifically designed for mass processing — for example, of incoming invoices.
Cryptalist Analysis: OpenAI's price dumping in the token segment reflects a strategic shift from maximizing margin to capturing market share ahead of Anthropic's public offering, which will set new valuation benchmarks for the entire AI sector. Investors should bear in mind that the price war is pressuring the profitability of both companies, while Anthropic's technological leadership in the intelligence ranking remains a significant argument in favor of its IPO narrative.