← All models
Balanced — strong quality at a mid priceOpen Weights
WizardLM-2 8x22B Pricing
Microsoft · microsoft-wizardlm-2-8x22b
Input / 1M tokens
$0.620
Output / 1M tokens
$0.620
Context window
65,535 tokens
≈ 87 pages of text
Max output
8k tokens
What it costs in practice
Typical request (1,200 in + 400 out tokens)$0.001
1,000 requests / month$0.992/mo10,000 requests / month$9.92/mo100,000 requests / month$99.20/moEstimate your own workloadOpens the calculator with WizardLM-2 8x22B preloaded — adjust volume and token counts there.
Price history
Only one price point recorded so far.
The staircase chart appears once a price change is detected.
What it excels at
Performs well on instruction following, multi-step reasoning, and coding tasks for its parameter scale.
The business tradeoff
Open-weights model; inference latency, throughput, and output consistency depend on the chosen hosting provider and quantization level.
Cheaper from Microsoft
Head-to-head
Vendor list rates, as of Jun 15, 2026 · source: openrouter · per-request examples assume 1,200 input + 400 output tokens.