Cumulus LabsBack to Leaderboard
LLMs
Qwen 3 14B
+14.7%~$QW3-14
14B reasoning model with extended thinking mode. Hybrid reasoning approach lets you trade compute for accuracy on demand. Runs on all 4 universal-a nodes.
Deploy Now
#10 overall · 4 deploys
1.4s
Cold Start
avg on Ion
310ms
Avg Inference
per request
4
Active Replicas
right now
4
Total Deployments
all time
7-Day Trend
+14.7% this weekCold Start vs Competitors
Lower is better●~$QW3-141.4s
~$QW3-81.2s
~$QW301.7s
~$QW321.8s
Cold start comparison vs similar models. Lower is better.
Cost Estimate
Cost per 1K Inferences
$0.85
Est. Daily (1K req/day)
$0.85
Est. Monthly (30K req)
$25.50
No subscriptions. Buy credits, pay per inference. Scale to zero when idle.
Quick Deploy
cumulus-sdkimport cumulus from "cumulus-sdk" // Deploy Qwen 3 14B on Ion const client = await cumulus.deploy("qwen3-14b") // Run inference const result = await client.run({ prompt: "Your prompt here", // model-specific params... })