OpenAI launches GPT Sol 6.1 with adaptive reasoning engine

OpenAI released GPT Sol 6.1, a model with adaptive test-time compute that scales reasoning depth to task complexity, answering Google's Gemini 4. It scores 81.4% Pass@1 on SWE-bench Verified, sustains 180 tokens/s and costs $0.30 per 1M input and $1.20 per 1M output tokens.

• 81.4% Pass@1 on SWE-bench Verified sets a new record
• Sustains 180 tokens per second in continuous generation
• Pricing: $0.30 per 1M input and $1.20 per 1M output tokens
• Time-to-first-reasoning-token under 92 ms