Compress tool outputs, logs, files, and RAG chunks before they reach the LLM. 20% fewer tokens for coding agents, 60-95% fewer tokens for JSON, same answers. Library, proxy, MCP server.
Python agentaianthropicclaude-codecompressioncontext-engineeringcontext-windowcursorfastapilangchainllmmcpopenaiprompt-engineeringproxypythonragtoken-optimizationtokenstypescript
Commits on main
deps: bump postcss from 8.5.19 to 8.5.25 in /plugins/opencode (#2748)
ff4e016 dependabot[bot] committed 3 months ago
deps: bump postcss from 8.5.19 to 8.5.25 in /plugins/openclaw (#2749)
cd60ee9 dependabot[bot] committed 3 months ago
deps: bump next from 16.2.10 to 16.3.0 in /docs (#2750)
0fd0b99 dependabot[bot] committed 3 months ago
deps: bump brace-expansion from 5.0.7 to 5.0.9 in /docs (#2751)
56ee57b dependabot[bot] committed 3 months ago
fix(litellm): add async_post_call_success_hook to HeadroomCallback (#1322)
3107994 Lucas Santos committed 3 months ago
fix: emit SSE ping before message_start on Bedrock streaming path (issue #902) (#1080)
4dab254 Focused Instability committed 3 months ago
docs(readme): surface the Serena opt-out in the wrap quickstart (#2790)
702a076 Tejas Chopra committed 3 months ago
docs(metrics): document OTLP metric export and Dynatrace ingest (#2785)
6b63b62 Tejas Chopra committed 3 months ago
ci(docs): deploy Pages on wiki changes, drop the never-working Vercel job (#2746)
a9a2fbd Tejas Chopra committed 3 months ago
fix(telemetry): stop mixing tokenizer scales in RequestOutcome, and fix the overhead framing (#2756)
04e1517 Tejas Chopra committed 3 months ago
refactor(pricing): make LiteLLM the source of truth, not the hardcoded table (#2779)
0e1d6bf Tejas Chopra committed 3 months ago
fix(router): stop counting an image's base64 payload as suffix tokens (#2778)
f03cc6d Tejas Chopra committed 3 months ago
fix(tokenizers): resolve gpt-5 and mixed-case model names to the right encoding (#2776)
fc4680b Tejas Chopra committed 3 months ago
fix(providers): stop a shorter model family shadowing a longer one (#2762)
0cb72f4 Tejas Chopra committed 3 months ago
fix(providers): give every model exactly one tokenizer (#2761)
cd92ed5 Tejas Chopra committed 3 months ago
fix(router): compare token quantities in one unit (#2759)
ad56dd3 Tejas Chopra committed 3 months ago
fix(cost): send litellm the total prompt so --budget stops seeing $0 (#2757)
a033ac4 Tejas Chopra committed 3 months ago
fix(providers): stop pricing modern content blocks at zero (#2760)
06add9e Tejas Chopra committed 3 months ago
fix(tokenizers): count HuggingFace chat templates, and resolve gpt-5 / gateway-wrapped names (#2758)
0ed306b Tejas Chopra committed 3 months ago
fix(savings): surface request growth the tok_saved clamp swallows (#2708)
184146b nangsontay committed 3 months ago
fix(compression): honor qualified CCR names across integrations (#2698)
dcb674b Rod Boev committed 3 months ago
fix(transforms): stop ContentRouter recompressing headroom_retrieve results (#2654)
677e097 michaeltarleton committed 3 months ago
fix(proxy): stop toggling headroom_retrieve in the Anthropic tools array (#2672)
08fce29 nangsontay committed 3 months ago
fix(deps): bump aiohttp and cryptography to clear the CVEs blocking 0.34.0 (#2753)
0221e7f Tejas Chopra committed 3 months ago
fix(compress): resolve the /v1/compress tokenizer per model, and document the real contract (#2743)
6422a80 Tejas Chopra committed 3 months ago
fix: normalize /p/<project> prefix on WebSocket upgrades so the Responses WS route is not rejected with 403 (#2379)
789a4f3 Matt Van Horn committed 3 months ago