Inference Mesh
Checking…
Clawd Inference Mesh
Live routing dashboard for the Clawd native model fleet. Both LoRA models run on https://clawd-inference-mesh.fly.dev — an Ollama-compatible backend on Fly.io. Every chat request tagged with a Clawd model ID routes here directly, bypassing OpenRouter entirely.
Mesh status
Checking…
No OLLAMA_BASE_URL configured
Native models
0
Core AI 1.5B + Trading Factory 8B
Mesh requests
…
Routed to clawd-inference-mesh this session
Total router requests
…
0 errors · 0% success
Live routing path
Browser/chat
/v1/chat
ZK Routerclawdrouter-zk
ollama route
Inference Meshfly.dev · Ollama
completion
Clawd Models1.5B · 8B LoRA
Clawd native models
2 models
1.5B LoRA
Free · Ollama
Solana Clawd Core AI
8bit/solana-clawd-core-ai:latestRoute
mesh → ollama
Cost
$0.00/M
Context
32K
8B LoRA
Free · Ollama
Solana Trading Factory 8B
8bit/solana-trading-factory:8b-lora-20260620Route
mesh → ollama
Cost
$0.00/M
Context
32K
Mesh configuration
OLLAMA_BASE_URLhttps://clawd-inference-mesh.fly.dev
OLLAMA_ENABLEDtrue
Core AI model8bit/solana-clawd-core-ai:latest
Trading model8bit/solana-trading-factory:8b-lora-20260620
Endpointhttps://clawd-inference-mesh.fly.dev/v1/chat/completions
Stats refresh every 5 s. Model routing is handled by the ZK Router at clawdrouter-zk.fly.dev — Clawd model IDs are detected and forwarded to the inference mesh automatically.