GPT-6.1 Terra (Max) hits #8 in Agent Arena, reshaping the Pareto frontier at $0.62/task
arena · x · 2026-10-29
Agent Arena (LMArena) published real-world results from nearly 9K agentic sessions: OpenAI's GPT-6.1 Terra (Max) ranks #8 with a +10.6% net improvement and a median cost of $0.62 per task — 43% cheaper than #7 Gemini 4 Argon (High) at $1.09 and 30% cheaper than GPT-6.1 Sol (Max) — reshaping the cost-quality Pareto frontier. The frontier chart shows Claude Opus 5.5 (Max) at the quality top (+21%), with MiMo-V2.6 variants in the low-cost region down to $0.01/task.
Related event: GPT-6.1 Terra Ranks 8th on Agent Arena at $0.62 Median Cost per Task(2 posts)→
More from coding & agent
- Personal agent digs through mail, calendar & Slack to find who to meet — signulll · 2026-10-30
- Inworld launches Reasoning for Realtime TTS-2, first context-steering TTS model — Scobleizer · 2026-10-30
- Relics packages domain-specific research machinery into installable agent swarm configs — Scobleizer · 2026-10-30
- Baseten partners with Postman to serve enterprise agents for 40M developers — baseten · 2026-10-30
- Pedro Domingos: Claude is 'pretty useless' for code never written before — pmddomingos · 2026-10-30
- AI Code Security Finally Improves: Veracode Benchmark Pass Rate Jumps From 56% to 62% — WeldPond · 2026-10-30