Solution · Cut agent spend # You pay for the same logs on every call. *Send the model less.* Why is your agent bill so high? It sends the same logs and files to the model on every call. Context Mode sends only what the model needs: one long session cost 65.5% less, $7.20 to $2.49. [Get started free](https://console.context-mode.com/login) [Read the docs](https://context-mode.com/docs/how-saving-works) Your agent Test output A 7 MB file Long session Your bill Your agent runs 401 tests. 1 fails. Printed **16.2 KB** Sent to the model **0.7 KB** Sent to the model - The failing test and its error - The summary line Kept in your archive - Every line, word for word Size measured on the minitest test fixtures, Claude Haiku 4.5. One example, measured by hand. How many versions of react are on npm? The script reads **7,011,585 B** The model reads **407 B** Sent to the model - The answer: 2,959 versions Never sent - The 7 MB file itself One query, not a benchmark figure. The benchmark result is under Your bill. One session, 30 steps. Tokens sent on each call. Without Context Mode **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 session, archive search on · Claude Opus 5.5 **$7.24 → $2.97** *59.0% lower* 4 of 4 runs](https://context-mode.com/docs/benchmarks) Here the agent can search its archive. Memory answers: 10, 10, 10 and 8 of 10. The same task, run twice. Cost per task. Without Context Mode, 30 steps **$7.20** Context Mode, 30 steps **$2.49** [Long session, archive search off · Claude Opus 5.5 **$7.20 → $2.49** *65.5% lower* 4 of 4 runs](https://context-mode.com/docs/benchmarks) [Data work on large files, as a script · Claude Haiku 4.5 **$0.175 → $0.062** *64% lower* 12 of 12 runs](https://context-mode.com/docs/benchmarks) [Test fix loop, minitest · Claude Opus 5.5 **$0.157 → $0.061** *61% lower* 8 of 8 runs](https://context-mode.com/docs/benchmarks) Archive search off in the 65.5% run. With it on: $7.24 to $2.97, 59.0% lower. Your own Anthropic or OpenAI account pays the model. We add no fee on tokens. The problem ## Your agent sends it all. *On every call.* **Tool output stays** A test run, a log or a JSON file enters the conversation and goes to the model again on every later call. **Big files, read in full** To answer one question about a large file, the agent reads it into the conversation, or sees only its first part. **Long sessions grow** Your agent sends the whole conversation on every call, and it grows with each step. How it works ## Keep your agent and your model. *Change what each call carries.* 1. 01 **Shorten long output** Context Mode shortens test, build and log output the first time it sees it. Errors stay in full. The rest goes to your archive. 2. 02 **Run data work as code** The agent writes a short script. Only what it prints reaches the model. 3. 03 **Archive older work** The model gets your recent work and an index of the rest. Older details are looked up word for word. **Your agent**Sends the full history and every tool output, as usual. → **Context Mode**Changes what the request carries. - Shortens long tool output - Runs data work as code - Moves older work to the archive → **The model**Reads a smaller request. Your own account pays for it. Your agent sends everything. Context Mode, in blue, makes the request smaller before the model reads it. The full text stays in your archive. Proof ## The same task, run twice. *Cost per task.* **65.5% lower** one long session on Claude Opus 5.5, $7.20 to $2.49, 4 of 4 pairs **64% lower** data work on large files on Claude Haiku 4.5, $0.175 to $0.062, 12 of 12 pairs **61% lower** a minitest fix loop on Claude Opus 5.5, $0.157 to $0.061, 8 of 8 pairs | Task | Context Mode | Without Context Mode | Lower | | --- | --- | --- | --- | | Long session, 30 steps, Opus 5.5 | $2.49 | $7.20 | 65.5% | | Data work on large files, as a script, Haiku 4.5 | $0.062 | $0.175 | 64% | | Test fix loop, minitest, Opus 5.5 | $0.061 | $0.157 | 61% | The 65.5% run had archive search off. With it on: $7.24 to $2.97, 59.0% lower, memory answers 10, 10, 10 and 8 of 10. On Opus 5.5 the same data task was 9 to 10% lower. Pre-registered paired runs at API list price, with task success equal in both arms. A short question with few tool calls leaves little to save; your own saving depends on your mix of work. [Method and sources →](https://context-mode.com/docs/benchmarks) Products used ## What this *uses.* **[Context Saving](https://context-mode.com/context-saving)** Pay for the failing test, not the ones that passed. **[Thinking in Code](https://context-mode.com/thinking-in-code)** Analyze a 7 MB file. Pay for 407 bytes. **[Recall](https://context-mode.com/recall)** All-day sessions with no compaction. Older work is looked up word for word. **[Active Context](https://context-mode.com/active-context)** Choose how much history the model sees: 200K, 500K or 1M. **[How saving works](https://context-mode.com/docs/how-saving-works)** Where an agent's tokens go, and what Context Mode shortens. **[Benchmarks](https://context-mode.com/docs/benchmarks)** Every paired run, with its evidence file. Related: [Long sessions](https://context-mode.com/solutions/long-sessions) · [For developers](https://context-mode.com/solutions/developers) · [For startups](https://context-mode.com/solutions/startups) · [Plans](https://context-mode.com/pricing) ## Keep your agent. *Pay for less of what it reads.* *npx @context-mode/cli* connects your agents. 1,000 requests free, no card. Your own Anthropic or OpenAI account pays the model. We add no fee on tokens. [Get started free →](https://console.context-mode.com/login) [Read the docs →](https://context-mode.com/docs/how-saving-works)