ai.arsentev/contextburn
contextburn
Run efficiency for coding agents: share of paid tokens that became output, not context re-reading.
alive probed 2026-10-19
2tools
≈ 158tokens, estimate
≈ <0.1 %of Claude's 200k window
0.1×the median server
327 msto initialize
3 msto list tools
- Run it
- uvx contextburn==0.2.1 (stdio, probed in a container with the network off)
- Host
- pypi.org
- Reports itself as
- contextburn 0.2.1
- Capabilities
- tools
- Instructions
- none
- Of a 1M window
- ≈ <0.1 % of a 1M context window, estimate
- Claude count
- The count_tokens endpoint refused the tool schemas: invalid x-api-key. The estimate stands in.
- Repository
- https://github.com/arsentev-ai/contextburn
- Website
- https://arsentev.ai
- Registry
- active, updated 2026-09-11 · registry entry
- Reachable
- 1 of the last 1 probe
Most expensive tools
Estimates with the o200k_base tokenizer.
| Tool | o200k estimate | Share of this server |
|---|---|---|
| run_efficiency | 89 | 56 % |
| spend_breakdown | 69 | 44 % |
Badge for the README
The result of the latest probe, the tool count and the token cost. The feed has an entry for every probe that changed them.
A mail when this server changes
Stops answering, comes back, tool set or cost changes. One confirmation mail first, a stop link in every mail.
For the publisher
- The Anthropic count_tokens endpoint refused the tool schemas (invalid x-api-key). A client that validates schemas can refuse them too.