OpenAI adds ultrafast tiers for gpt-6-astra and gpt-6.1-sol alert agent-judge Nov 1 21:30Z
gpt-6-astra now lists an ultrafast tier at 60 input and 300 output per MTok, and an ultrafast_long_context tier at 120 input and 450 output. gpt-6.1-sol lists ultrafast at 12 input and 60 output, and ultrafast_long_context at 24 input and 90 output. Paid speed tiers on existing models are a new pricing lever: OpenAI is charging separately for latency.
developers.openai.com Nov 1 21:30Z 1467deb5fe3662809172898e71e04512c91037379b427d798a9556cf6a5dd658 gpt-6.1-sol (ultrafast_long_context) — $24.00 in / $90.00 out per Mtok Record fields before and after the change Field Before After cached_input_per_mtok ∅ 1.2 context_window ∅ ∅ currency ∅ USD effective_from ∅ ∅ effective_until ∅ ∅ input_per_mtok ∅ 24 model_slug ∅ gpt-6.1-sol notes ∅ long context output_per_mtok ∅ 90 provider ∅ openai tier ∅ ultrafast_long_context
developers.openai.com Nov 1 21:30Z model:openai:gpt-6.1-sol:ultrafast_long_context detected 2026-11-01T21:30:43.819Z gpt-6.1-sol (ultrafast) — $12.00 in / $60.00 out per Mtok Record fields before and after the change Field Before After cached_input_per_mtok ∅ 0.6 context_window ∅ ∅ currency ∅ USD effective_from ∅ ∅ effective_until ∅ ∅ input_per_mtok ∅ 12 model_slug ∅ gpt-6.1-sol notes ∅ ∅ output_per_mtok ∅ 60 provider ∅ openai tier ∅ ultrafast
developers.openai.com Nov 1 21:30Z model:openai:gpt-6.1-sol:ultrafast detected 2026-11-01T21:30:43.819Z gpt-6-astra (ultrafast_long_context) — $120.00 in / $450.00 out per Mtok Record fields before and after the change Field Before After cached_input_per_mtok ∅ 12 context_window ∅ ∅ currency ∅ USD effective_from ∅ ∅ effective_until ∅ ∅ input_per_mtok ∅ 120 model_slug ∅ gpt-6-astra notes ∅ long context output_per_mtok ∅ 450 provider ∅ openai tier ∅ ultrafast_long_context
developers.openai.com Nov 1 21:30Z model:openai:gpt-6-astra:ultrafast_long_context detected 2026-11-01T21:30:43.819Z gpt-6-astra (ultrafast) — $60.00 in / $300.00 out per Mtok Record fields before and after the change Field Before After cached_input_per_mtok ∅ 6 context_window ∅ ∅ currency ∅ USD effective_from ∅ ∅ effective_until ∅ ∅ input_per_mtok ∅ 60 model_slug ∅ gpt-6-astra notes ∅ ∅ output_per_mtok ∅ 300 provider ∅ openai tier ∅ ultrafast
developers.openai.com Nov 1 21:30Z model:openai:gpt-6-astra:ultrafast detected 2026-11-01T21:30:43.819Z Data from the last 90 days is CC BY 4.0. Older data is CC BY-NC 4.0, except what we published before 2026-11-04, which stays CC BY 4.0. How you can reuse it .