GPT-6.1 Terra Questioned as Benchmark Ties with Sol but Costs More
GPT-6.1 Terra tied GPT-6 Astra at 57.6% on Agents' Last Exam but cost 16.6x more per task, while Reddit benchmarks and cost analyses show it loses to the cheaper sibling GPT-6.1 Sol, which ends up less expensive in practice despite Terra's lower token pricing.
2026-10-27 ~ 2026-10-27 · 3 related posts
- Episode 1: Microsoft PR leak reveals unreleased OpenAI model GPT-6.1 Terra(2026-10-19, 6 posts)
- Episode 2: Suspected OpenAI model GPT-6.1 Terra surfaces across multiple channels ahead of likely launch(2026-10-25, 10 posts)
- Episode 3: Report: OpenAI pauses frontier RL training, shifts to smaller cheaper models(2026-10-25, 2 posts)
- Episode 4: OpenAI launches GPT-6.1 Terra: fast, cheap computer-use model(2026-10-27, 46 posts)
- Episode 5: GPT-6.1 Terra Slashes Factual Error Rate from 22.3% to 4.3%(2026-10-27, 3 posts)
- Episode 6: Artificial Analysis Rolls Out Intelligence Index v4.3.2 with GPT-6.1 Terra(2026-10-27, 3 posts)
- Episode 7: OpenAI Quietly Swaps Pro Default Model and Halves Usage Quota(2026-10-27, 2 posts)
- Episode 8: GPT-6.1 Terra Questioned as Benchmark Ties with Sol but Costs More(2026-10-27, 3 posts)
- Real-world math: GPT-6.1 Terra's half price is offset by 2-3x token usage, costing more than Sol — lordpuddingcup · 2026-10-27
- GPT-6.1 Terra matches GPT-6 Astra at 57.6% on Agents' Last Exam at 1/16.6x cost — aakashgupta · 2026-10-27
- GPT 6.1 Sol outperforms GPT 6.1 Terra at the same cost in nearly every benchmark — Lucky_Creme_5208 · 2026-10-27