OpenAI: GPT-6.1 Sol Pro

GPT-6.1 Sol Pro is OpenAI's high-compute execution mode for GPT-6.1 Sol, exposed on the gpt-6.1-sol API by setting reasoning.mode to pro. It uses the same proprietary weights as the standard Sol snapshot but spends far more reasoning tokens per request, trading latency and cost for stronger performance on difficult coding, agentic, multimodal, and professional workflows.

The model ships with a 1,050,000-token context window, up to 128,000 output tokens, text and image (plus document/file) inputs, tool calling, structured outputs, and web search integration. It was released on September 29, 2026 as an upgrade to GPT-6 Sol, positioned between the lower-cost GPT-6 Luna tier and flagship GPT-6 Astra for teams that want near-Astra agentic capability at roughly one-fifth of Astra's standard API token prices.

Public evaluations highlight the Sol family's strength on real repository engineering and business automation: OpenAI reports 75.2% on DeepSWE v1.1 and 31.7% on AutomationBench 1.0.6 at published reasoning settings, while independent leaderboards at max or Pro-oriented settings show competitive GPQA Diamond, MMLU-Pro, MMMU-Pro, LiveBench, SWE-Bench, and chat-arena Elo results. Pro mode is intended for high-stakes tasks where extra reasoning depth justifies several times higher per-request cost versus standard GPT-6.1 Sol.