Shopify gist tokens compress a 4,200-token system prompt into 47 learned tokens with 95% task…
shopify gist tokens compress a 4,200-token system prompt into 47 learned tokens with 95% task accuracy. that is not prompt engineering anymore — that is a new file format for prompts.
everyone shipping long system prompts in 2027 will look like everyone shipping minified-by-hand css in 2014.
Context
Shopify Engineering's post of 19 August 2026 describes gisting, which uses a set of learned tokens. It reports the Sidekick GraphQL agent system prompt going from about 6,000 tokens to about 1,500 gist tokens (4:1) without losing prediction quality, with serving gains at 350 requests per second: time to first token from 438 ms to 354 ms, end-to-end from 6.8 s to 4.2 s, and throughput from 20.2 to 23.4 queries per second.
The note's figures of a 4,200-token prompt, 47 learned tokens and 95% task accuracy are unverified: none appear in the Shopify text read, which gives 6,000 to 1,500 instead. That does not show them false, since the note may draw on another source. Gisting setups can only be compared within their actual scope. The post predates the note. The new file format for prompts and the 2027 comparison are the author's opinion.
Related work
- Gisting write-up (InfoQ) ↗Secondary; snippet repeats the 6,000 and 1,500 figures.
- Gisting entry (ZenML LLMOps database) ↗Secondary; snippet only.
Watch next
- The source of the 4,200, 47 and 95 figures, if any.
Sources
- Gisting (Shopify Engineering, 19 Aug 2026)shopify.engineering
Provenance
The note above is reproduced unedited from the original post, first published on Threads on 5 September 2026 at 10:02 IST. Sources are the papers and datasets the note draws on.
View the original post ↗Embed this note
More notes
The air is now being asked to keep its own ledger
the air is now being asked to keep its own ledger: ecmwf’s aifs compo becomes the first ai model to forecast atmospheric composition globally every three hours, cleanair simulates 365 days of pm2.5 over china in ten seconds, and a unified framework maps six pollutants at one kilometer across the whole country. the air now files its own composition report.
read the note →The current is now being asked to draw its own map
the current is now being asked to draw its own map: china’s langya 2.0 predicts six ocean phenomena including internal waves and mesoscale eddies, a deep net called wenhai resolves eddies globally with air sea flux formulas built in, and scripps infers surface currents from the way temperature patterns deform in satellite images. the ocean now files its own circulation report.
read the note →The soil is now being asked to report its own carbon
the soil is now being asked to report its own carbon: a nix color sensor paired with generative data augmentation predicts soil organic carbon without a lab, random forest drives 74 percent of soil health mapping studies, and sentinel 2 tracks five year carbon change across france and italy from 922 samples. the dirt now files its own carbon account.
read the note →