Qwen3 14B vs WizardLM-2 8x22B.
model vs model · prices as of
Qwen3 14B takes 4 of 5 rounds against WizardLM-2 8x22B. It is also the cheaper of the two to run, at $0.41 per million tokens against $1.92.
Alibaba Cloudwins 4 roundsQwen3 14B
$0.41/M tokens
cheapest we can explain, at Nscale · $0.054 per 1,000 pages
131k context · tool calling · open weights
WizardLM-2 8x22B
$1.92/M tokens
cheapest we can explain, at DeepInfra · $0.317 per 1,000 pages
66k context · open weights
List price in (lower wins)
$0.35/M
$0.62/M
List price out (lower wins)
$1.40/M
$0.62/M
Cheapest we can explain (lower wins)
$0.41/M at Nscale
$1.92/M at DeepInfra
Standard price (lower wins)
$2.45/M
$2.48/M
Cost per 1,000 pages (lower wins)
$0.054
$0.317
Context
131k tokens
66k tokens
Open weights
yes
yes
Tool calling
yes
no
Structured output
not stated
yes
Reads images
no
no
Reasoning
yes
no
Released
1 Apr 2025
16 Apr 2024
Providers selling it
19
8
Price we measure against
the maker's own
OpenRouter
Trains on your prompts (no is better)
not stated
not stated
Providers selling both · 6
NanoGPT$0.48 / $1.97
DeepInfra$0.60 / $1.92
Eden AI$0.60 / $1.92
TrustedRouter$0.43 / $2.62
OpenRouter$0.60 / $2.48
Kilo Gateway$1.59 / $2.48
Qwen3 14B then WizardLM-2 8x22B, per million tokens
Go further
Add a third or fourth model, or switch the weighting, in the tool.
What moved in the last seven days
Frequently asked
- Which is cheaper, Qwen3 14B or WizardLM-2 8x22B?
- Qwen3 14B, at $0.41 per million tokens against $1.92 for WizardLM-2 8x22B, three parts input to one part output, read on 16 Sept 2026.
- Where is each one cheapest?
- Qwen3 14B: $0.41 at Nscale. WizardLM-2 8x22B: $1.92 at DeepInfra.
- Can I buy both from one provider?
- Yes. 6 providers sell both: NanoGPT, DeepInfra, Eden AI, TrustedRouter, OpenRouter, Kilo Gateway. The cheapest of them for the two together is NanoGPT.
- Which holds more context?
- Qwen3 14B, at 131k tokens against 66k.
Prices exclude tax and are per million tokens, read 16 Sept 2026 from each provider's own catalog. The standard price blends three parts input to one part output; a page is 600 tokens in and 60 out. Bars are scaled to the larger value in each row.