📡 Astra Telemetry Service v2

Edge cache â€ĸ ● OPTIMIZED ⚡ 1.5s → 5ms cached â€ĸ Stealth AI Gateway

kkr-astra-cache.iamkkronly58.workers.dev astra-telemetry-service.hf.space v2-cached

🚀 Speed Fix - āĻāĻ–āύ āĻ…āύ⧇āĻ• āĻĻā§āϰ⧁āϤ!

āφāϗ⧇āĻāĻ–āύāĻ•āĻŋāĻ­āĻžāĻŦ⧇
5-8 āϏ⧇āϕ⧇āĻ¨ā§āĻĄ1.4-1.6s (āĻĒā§āϰāĻĨāĻŽāĻŦāĻžāϰ)1B āĻŽāĻĄā§‡āϞ + max_tokens 150
āĻĒā§āϰāϤāĻŋāĻŦāĻžāϰ 5s5-24ms (cached)Cloudflare Edge Cache 60s
512 tokens150 tokens defaultāĻĻā§āϰ⧁āϤ āĻœā§‡āύāĻžāϰ⧇āĻļāύ

Same query 60 āϏ⧇āϕ⧇āĻ¨ā§āĻĄā§‡āϰ āĻŽāĻ§ā§āϝ⧇ āφāĻŦāĻžāϰ āĻ•āϰāϞ⧇ X-Cache: HIT āĻāĻŦāĻ‚ 5ms āĻ response! quotes āϏāĻŦ double quotes āχ āϰāĻžāĻ–āĻž āĻšāϝāĻŧ⧇āϛ⧇āĨ¤

🔗 Live API (quotes same)

https://kkr-astra-cache.iamkkronly58.workers.dev/api?chat=hello%20ai
→ GET /api?chat=hello%20ai (5ms cached)
→ Uncensored Stheno 8B
→ /health (memory usage)

📤 Response Format (double quotes)

{
  "status": "ok",
  "query": "hello ai",
  "response": "How can I assist you today?",
  "model": "meta-llama/Llama-3.2-1B-Instruct",
  "latency_ms": 5,
  "cached": true,
  "developer": "Kaustav Kanti Ray @iamkkronly"
}

🤖 āφāϰāĻ“ Uncensored āϛ⧋āϟ āĻŽāĻĄā§‡āϞ āϝ⧋āĻ— āĻ•āϰāĻž āĻšāϝāĻŧ⧇āϛ⧇

Short CodeModel IDSizeSpeedāϧāϰāύ
llama-1bmeta-llama/Llama-3.2-1B-Instruct1B1.5s ⚡Fastest default
llama-3bmeta-llama/Llama-3.2-3B-Instruct3B3.9sPermissive
stheno-8bSao10K/L3-8B-Stheno-v3.28B1.7s đŸ”ĨUncensored
hermes-8bNousResearch/Hermes-3-Llama-3.1-8B8B5.7sUncensored Hermes
dolphin-24bCognitiveComputations/dolphin-mistral-24b-venice-edition24B1.7sUncensored Dolphin
gemma-4bgoogle/gemma-3-4b-it4B5.3sBalanced
qwen-4bQwen/Qwen3-4B-Instruct-25074B5.2sQwen best

āĻŦā§āϝāĻŦāĻšāĻžāϰ: /api?chat=hello&model=stheno-8b āĻŦāĻž &model=dolphin-24b — āϏāĻŦ double quotes āĻ JSON āϰāĻŋāϟāĻžāĻ°ā§āύ āĻšāĻŦ⧇āĨ¤

💾 Space & Memory āĻŦā§āϝāĻŦāĻšāĻžāϰ (Bangla)

āϜāĻžāϝāĻŧāĻ—āĻžāĻŦā§āϝāĻŦāĻšāĻžāϰāϞāĻŋāĻŽāĻŋāϟāĻŦāĻžāĻ•āĻŋ
HF Space (static)6.98 KB (5 files)300 MB99.9% āĻĢāĻžāρāĻ•āĻž
├─ index.html4.9 KB--
├─ README.md0.17 KB--
├─ api.html0.14 KB--
Cloudflare Worker10.1 KB script1 MB99% āĻĢāĻžāρāĻ•āĻž
├─ Memory~5-10 MB used128 MB92% āĻĢāĻžāρāĻ•āĻž
├─ CPU5-20ms per req50ms60% āĻĢāĻžāρāĻ•āĻž
├─ Cache60s edge cache-HIT = 5ms
└─ HardwareStatic (no GPU, no RAM) - āϏāĻŽā§āĻĒā§‚āĻ°ā§āĻŖ āĻĢā§āϰāĻŋ

💡 HF Space static āĻšāĻ“āϝāĻŧāĻžāϝāĻŧ āϕ⧋āύ⧋ GPU/RAM āϞāĻžāϗ⧇ āύāĻž, Cloudflare Worker edge āĻ āϚāϞ⧇ — āϤāĻžāχ āĻāϤ āĻ•āĻŽ memory āĻŦā§āϝāĻŦāĻšāĻžāϰāĨ¤