summarise a 100-page report50,000 tokens in, 1,500 out, one request
$0.12
$0.12
1,000 classification calls500 tokens in, 20 out, per call
$1.20
$1.20
coding-agent session2M tokens in across ~60 turns, 60,000 out, no cache
$4.60
$4.60
same session, cached85% of input read from the prompt cache, the rest written to itGPT-6 Sol costs 1.1× as much as GPT-6.1 Sol
$1.52
$1.69
at listed rates. token counts are typical sizes, not measurements.
GPT-6.1 Sol pricing notes
requests over 272K input tokens bill the entire request at 2x the input and cache rates and 1.5x the output rate; the listed price is the short-context rate.
cache hits bill at $0.10 per 1M tokens, half GPT-6 Sol's $0.20; cache writes cost $2.50 (30-minute) or $5.00 (12-hour).
GPT-6 Sol pricing notes
above 272K input tokens the whole request bills at long-context rates: 2x input and cache, 1.5x output.
OpenAI's GPT-6.1 Sol, for complex coding and professional work, replaced GPT-6 Sol at the same $2 / $10 rate with half Sol's cache-hit price. it sits below GPT-6 Astra and above GPT-6 Luna in price.
pick it if
you want complex coding work at one-fifth of GPT-6 Astra's token price
you resend long prompts; cache hits bill at 5% of input
your prompts mix text and images
skip it if
GPT-6 Luna at $0.10 / $0.50 covers your workload
your prompts often pass 272K input tokens; the whole request then bills higher
GPT-6 Sol was OpenAI's mid-tier GPT-6 model, between Astra and Luna, taking text and image input with a reasoning mode. GPT-6.1 Sol replaced it at the same $2 / $10 with cheaper cache reads.
pick it if
you needed GPT-6 reasoning at a fifth of Astra's price
you wanted a 1M-token window with text and image input
you reused long prompts: cache reads billed at a tenth of input
skip it if
you are choosing today; GPT-6.1 Sol halves cache reads at the same price
faq
which is cheaper, GPT-6.1 Sol or GPT-6 Sol?
GPT-6.1 Sol and GPT-6 Sol list the same input and output prices. on a coding-agent session (2M tokens in, 60,000 out) GPT-6.1 Sol costs $4.60 and GPT-6 Sol $4.60.
try either model
switch between models inside the same octomind session.