Context Saving

# Pay for the failing test, *not the 599 that passed.*

Context Saving folds long tool output before the model reads it. Errors stay in full. The rest is archived, one lookup away.

[Get started free](https://console.context-mode.com/login) [Read the docs](https://contextmode.ai/docs/how-saving-works)

Try it

Test-Driven Debugging Large Tool-Output Handling Long Agentic Coding Session

Your agent runs 600 tests. 1 fails.

Printed **17.0 KB** Sent to the model **1.2 KB**

Sent to the model

- The failing test and its error
- The summary line
- One line for the 599 that passed

Kept in your archive

- All 600 results, word for word

[npm test fix loop · Claude Opus 5.5 **$0.087 → $0.068** *22% lower* 15 of 16 runs](https://contextmode.ai/docs/benchmarks)

Your agent reads an API response with 800 orders.

Printed **800 orders** Sent to the model **4 orders**

Sent to the model

- Every field name
- The first 3 orders and the last
- A line that counts the rest

Kept in your archive

- The other 796 orders

[One large JSON file · Claude Opus 5.5 **$0.086 → $0.068** *21% lower* 12 of 12 runs](https://contextmode.ai/docs/benchmarks)

One session, 30 steps. Tokens sent on each call.

Direct **330,785** Context Mode **75,591**

Sent to the model

- Your recent work, unchanged
- An index of older work

Kept in your archive

- Older work, word for word

[Long agentic coding session · Claude Opus 5.5 **$7.20 → $2.49** *65.5% lower* 4 of 4 runs](https://contextmode.ai/docs/benchmarks)

How it works

## 17 KB printed. *1.2 KB read.*

1. 01 **Your agent runs a command** Tests, builds, logs, big JSON.
2. 02 **Context Mode folds the output** Errors and summaries stay as printed.
3. 03 **The model reads less** And can read the full output back by id.

Proof

## The same task, run twice. *Cost per task.*

Test fix loop, minitest

Direct **$0.157** Context Mode **$0.061**

61% lower

npm test fix loop

Direct **$0.087** Context Mode **$0.068**

22% lower

One large log

Direct **$0.091** Context Mode **$0.071**

22% lower

One large JSON file

Direct **$0.086** Context Mode **$0.068**

21% lower

Paired runs, Claude Opus 5.5, at API list price. [Method and sources →](https://contextmode.ai/docs/benchmarks)

Features

## Folded the first time the model sees it.

**15 test runners**

Failures stay, passing tests become one line.

**JSON**

Every key stays. Long arrays keep 3 items and the last.

**Builds and logs**

Your stack frames stay; library frames fold.

**Cache-safe**

A result changes once, so later requests reuse the cache.

Use cases: [Failing test output](https://contextmode.ai/use-cases/failing-test-output) · [Keep the prompt cache warm](https://contextmode.ai/use-cases/prompt-cache)

## Try it on your *next test run.*

*npx @context-mode/cli* connects your agents. 1,000 requests free, no card.

[Get started free →](https://console.context-mode.com/login) [Read the docs →](https://contextmode.ai/docs/how-saving-works)