v3.2.0

Released September 8, 2026 · View on GitHub →

TL;DR — GPT-6 Astra and Gemini 3.8 Flash are in the model picker. Astra is not the default: it costs twice what GPT-5.6 Sol does and is still rolling out by organization.

Added

September. 1M context, 128K output, and /thinking works on it.

Not the default, on purpose. It is $10/$50 per million tokens against 5.6 Sol's $5/$30 — twice the price — and OpenAI is still rolling it out organization by organization, so a key that works today may not have it. GPT-5.6 Sol stays the default; Astra is one /model away.

September, at the same price as 3.7 Flash ($0.75/$3.75 promotional). 1M context, 64K output, tuned for long-horizon agentic work. If you are on 3.7 Flash there is no cost reason not to move.

Existing model ids are untouched. New ones are added beside, never in place of — dropping an id does not leave a pinned config on the older model, it drops the lookup and lands that config on another provider's default.

Fixed

gpt-5 prefix, so a reasoning model would have arrived without it and with no error to explain why.

temperature, top_p and top_k in the 3.7 generation and 3.8 is built on 3.7, so it inherits the removal — the parameters are now withheld for it too.

Cost estimates use the promotional Gemini rate that runs to 2026-12-31. The rise scheduled for 2027-01-01 is deliberately not written ahead of time: doing exactly that for Sonnet 5 over-reported every run by 50% until it was caught by hand. Cached reads on GPT-6 Astra are $1.00/M — 0.1x input, which the default rate already applies.

Downloads

Install with npm install -g codeep@3.2.0.