| Group | Cold p50 |
|---|---|
| Before | 4,140 ms |
| After | 1,644 ms |
How we cut cold search from 4.1 s to 1.6 s
The ContextFetch team
ContextFetch blog
Featured
How we cut cold search from 4.1 s to 1.6 s
Parallel recall, a fetch phase with a deadline, one batched rerank and two caches. And the part of the speed-up that was not engineering.
The ContextFetch team
| Group | Cold p50 |
|---|---|
| Before | 4,140 ms |
| After | 1,644 ms |
| Group | Share of cases |
|---|---|
| Answer in top highlight | 86.2% |
| Random chunk | 18.0% |
Benchmark methodology: what we measure and what we don't
The ContextFetch team
HTTP/2 402 x-payment-url: https://api.contextfetch.dev/buy/instant {"error": "keyless_limit", "action": {"type": "payment_link", ...}, "packs": [...]}
Keyless MCP and the 402 payment loop for agents
The ContextFetch team