What it costs to run GLM-4.5-Air (MoE)

How these numbers are produced

• Model shape comes from the model's own config.json — 46 layers, 96 attention heads, 8 KV heads.
• Weight size is parameters × effective bits-per-weight. Those constants are checked against real published quantised file sizes — median error 0.7% across 40 measurements.
• KV cache is 2 × layers × KV-heads × head-dim × context × 2 bytes. Using KV-heads rather than attention heads is what makes this correct for grouped-query attention; treating a GQA model as multi-head overstates the cache by up to 8×.
• Tokens/sec is an estimate, not a benchmark. Generation is memory-bandwidth-bound, so this is bandwidth ÷ weight-bytes derated to 75%. Real throughput depends on your runtime, batch size and kernels. We don't run our own hardware tests — see our editorial policy.
• This is a mixture-of-experts model. Generation reads only the experts active for each token (8 of 128), so throughput is scored on active parameters, not the full 106B. Memory is the opposite: every expert must still be resident in VRAM, so the figures above use the full weight set.Derived from architecture (config.json); MoE routes ~12B active params per token.
• Some memory-bandwidth figures are not yet independently verified.Where that's the case we show no tokens/sec at all rather than a number we can't stand behind.
• Rental prices are pulled hourly from provider APIs (last updated Tue, 13 Oct 2026 01:20:22 GMT). Spot/interruptible pricing can change or vanish without notice.