Infrastructure for the new open stack

Price a GLM workload before you rent

Start with 3B-active GLM-4.7-Flash, then compare the hardware behind Kimi, Qwen, MiniMax, StepFun, DeepSeek, and the rest of today's sparse, agentic stack.

See the cheapest practical setup Check RunPod availability Browse the new open stack
  • GLM-4.7-Flash
  • Interactive inference
  • No account or API key
Keep GPU prices on your radar Read the weekly brief Follow via RSS
Kimi K2.7 GLM-5.2 Qwen3.6 MiniMax M3 Step 3.7
Live starting point Practical GLM inference $0.44/hr A40 on RunPod Updated Jul 27, 2026
Model
GLM-4.7-Flash
Traffic
Interactive
GPU class
32GB+
Check the full recommendation
Need a different workload? Start from a current price floor.
Kimi long context What is the current 48GB+ price floor? $0.44/hr · A40 on RunPod Updated Jul 27, 2026 Heavy sparse inference What does a 140GB+ GPU cost right now? $2.29/hr · H200 on Lambda Updated Jul 27, 2026 Public-cloud Qwen What is the lowest public-cloud 48GB+ row? $3.09/hr · A100 SXM4 on AWS Updated Jul 27, 2026

Cross-provider comparison

Auditable rows Every visible price carries source, freshness, region, and offer-count context. Marketplace aggregates, public list prices, and provider API rows are labeled so missing or cheap cells are not over-read.
Methodology Freshness
Loading pricing data...

High-intent GPU guides

Cheapest GPU for GLM-4.7-Flash → Best GPU for Kimi Linear 48B hosting → Best GPU for Qwen3.6 35B-A3B hosting → Best cloud GPU for fine-tuning → Cheapest provider for batch inference → H100 vs A100 for training cost →

Popular provider comparisons

Vast.ai vs RunPod → AWS vs GCP → AWS vs Azure → Lambda vs RunPod → Vast.ai vs Lambda → GCP vs Azure →

Browse GPU pricing

Popular GPU comparisons

H100 SXM vs H200 → H100 SXM vs A100 SXM4 → RTX 4090 vs RTX 5090 → H100 SXM vs B200 → A100 SXM4 vs A100 PCIE → L40 vs RTX 6000Ada →
Agent-assisted setup

Hand this runbook to Claude Code or Codex

Open a terminal in the repository where you want the deployment files, start claude or codex, then paste this prompt. It asks the agent to verify sources and stop before it creates billable infrastructure.

Claude Code quickstart ↗ Codex docs ↗
Deployment prompt 
Use the infrastructure or model context on this page to create a reproducible open-model deployment. Use this guide as the starting context: https://www.getflops.ai/gpu-prices. Read the linked model card and provider documentation before choosing hardware or runtime settings. Open every linked primary source and flag any mismatch instead of guessing. Create a deployment folder containing README.md, .env.example with no secrets, a pinned start script or infrastructure manifest, and smoke-test.sh. Make the endpoint OpenAI-compatible where the runtime supports it. Run local/static validation, estimate the billable resources, and stop before provisioning paid infrastructure until I approve.
Guardrails included No secrets in files · verify primary docs · approval before spend