BroadRouter

Llama 3.2 3B vs Qwen3-14B API pricing

Updated

Llama 3.2 3B: USD 0.0509 input, USD 0.335 output per 1M tokens (Featherless). Qwen3-14B: USD 0.07 input, USD 0.28 output per 1M tokens (SiliconFlow). At 3 input tokens per output token, Qwen3-14B costs 0.5% more than Llama 3.2 3B (USD 0.1225 vs USD 0.1219 per 1M).

Picked by the same-tier rule: both are in the first quarter of the indexed models by price (below USD 0.1556 per 1M tokens at 3:1) and in the same size class, small (a tier word such as Flash, Mini or Haiku, or 40B parameters or fewer). Qwen3-14B is the 2nd-closest-priced model from another maker to Llama 3.2 3B in that bucket. Price only: every number is computed from prices read from each provider's own pricing page (linked per cell, newest check 2026-11-15), with no claim about quality, speed or accuracy.

How do Llama 3.2 3B and Qwen3-14B differ on price?

Llama 3.2 3B and Qwen3-14B prices side by side

USD per 1M tokens, standard tier. "Cheapest" is the lowest price at 3 input tokens to 1 output token; each line can come from a different provider.

Llama 3.2 3B and Qwen3-14B API prices compared
PriceLlama 3.2 3BQwen3-14B
Model pageLlama 3.2 3B Instruct
Meta · 4 providers
Qwen3-14B
Qwen · 3 providers
Cheapest standard price, blended 3:1USD 0.1219 lower
USD 0.0509 in / USD 0.335 out at Featherless
Featherless page checked 2026-11-15
USD 0.1225
USD 0.07 in / USD 0.28 out at SiliconFlow
SiliconFlow page checked 2026-11-14
Cheapest inputUSD 0.0509 lower
Featherless source
USD 0.07
SiliconFlow source
Cheapest cached inputNo standard price lists a cache price; cached tokens bill at the input price.No standard price lists a cache price; cached tokens bill at the input price.
Cheapest outputUSD 0.15 lower
Amazon Bedrock source
USD 0.24
DeepInfra source
Maker's own APIMeta publishes no paid standard price on its own API that we found.Qwen publishes no paid standard price on its own API that we found.
Median standard priceUSD 0.1005 / USD 0.335
input / output, 4 prices
USD 0.12 / USD 0.24
input / output, 3 prices
Context window128,000 tokens, largest stated; from 32,768 at some providers131,072 tokens, largest stated; from 32,768 at some providers
Providers with a standard price43

Report a wrong price or number (email to info@broadnet.ai, with this page's link filled in).

What do Llama 3.2 3B and Qwen3-14B cost for a real workload?

30 days at each model's cheapest standard price for the workload (presets of the cost calculator; cached input at the cached price where listed).

Monthly cost of three workloads on Llama 3.2 3B and Qwen3-14B
WorkloadLlama 3.2 3BQwen3-14BDifference
Support chatbot
2,000 requests a day, 1,500 input tokens (60% cached), 300 output
USD 10.61
Featherless
USD 11.34
SiliconFlow
USD 0.73
Llama 3.2 3B lower
RAG over documents
1,000 requests a day, 8,000 input tokens (30% cached), 500 output
USD 17.24
Featherless
USD 21.00
SiliconFlow
USD 3.76
Llama 3.2 3B lower
Coding agent (one developer)
1,500 requests a day, 40,000 input tokens (85% cached), 2,000 output incl. reasoning
USD 121.95
Cloudflare Workers AI
USD 151.20
SiliconFlow
USD 29.25
Llama 3.2 3B lower

Llama 3.2 3B is cheaper on all three workloads. The widest gap is on Coding agent (one developer): the dearer model costs 24% more than the cheaper one.

Does the cheaper model change with your input:output ratio?

The cheaper model changes with the mix: below 1.6 input tokens per output token, Llama 3.2 3B is cheaper; from 1.6 to 2.9, Qwen3-14B is cheaper; above 2.9, Llama 3.2 3B is cheaper. At the site's 3:1 blend, Llama 3.2 3B is cheaper. More input per output token, as with retrieval or agents that resend long contexts, favours the lower input price; long answers favour the lower output price.

Price per 1M tokens of Llama 3.2 3B and Qwen3-14B at different input:output mixes
Input : output tokensLlama 3.2 3BQwen3-14B
1 : 3 (long answers)USD 0.15 lower
Amazon Bedrock
USD 0.21
DeepInfra
1 : 1USD 0.15 lower
Amazon Bedrock
USD 0.175
SiliconFlow
3 : 1 (the site’s blend)USD 0.1219 lower
Featherless
USD 0.1225
SiliconFlow
10 : 1 (retrieval, long prompts)USD 0.07673 lower
Featherless
USD 0.08909
SiliconFlow
30 : 1 (agents, big contexts)USD 0.06006 lower
Featherless
USD 0.07677
SiliconFlow

USD per 1M tokens at the cheapest standard provider for each mix, before cache discounts. Thinking tokens bill as output.

What do Llama 3.2 3B and Qwen3-14B cost per character in other languages?

Cost to send 1M characters at each model's cheapest non-promotional standard input price, with the Llama 3.3 70B tokenizer for Llama 3.2 3B and the Qwen3 32B tokenizer for Qwen3-14B:

Input cost per 1M characters by language, Llama 3.2 3B and Qwen3-14B
ContentLlama 3.2 3BQwen3-14B
EnglishUSD 0.011 lower
Featherless
USD 0.015
SiliconFlow
Arabic MSAUSD 0.020 lower
Featherless
USD 0.027
SiliconFlow
Egyptian ArabicUSD 0.020 lower
Featherless
USD 0.028
SiliconFlow
HindiUSD 0.027 lower
Featherless
USD 0.065
SiliconFlow
ChineseUSD 0.041 lower
Featherless
USD 0.045
SiliconFlow
SpanishUSD 0.014 lower
Featherless
USD 0.019
SiliconFlow
CodeUSD 0.013 lower
Featherless
USD 0.018
SiliconFlow

From the real cost by language dataset (measured tokens per character on FLORES-200 and other texts). Your own text will differ.

Check it on your own traffic

More comparisons: Llama 3.2 3B vs Llama 3.2 1B · Gemma 4 26B A4B vs Qwen3-14B · Qwen3 30B A3B vs Qwen3-14B · Qwen3-14B vs Qwen3-8B · every Llama 3.2 3B comparison · every Qwen3-14B comparison · all 662 price comparisons

Questions

Is Llama 3.2 3B cheaper than Qwen3-14B?

Yes. On a blend of 3 input tokens to 1 output token, Llama 3.2 3B costs USD 0.1219 per 1M tokens and Qwen3-14B USD 0.1225, at each model's cheapest standard price. For input alone, Llama 3.2 3B is cheaper (USD 0.0509 against USD 0.07 per 1M); for output alone, Llama 3.2 3B is (USD 0.15 against USD 0.24). This compares price only, not quality.

What do Llama 3.2 3B and Qwen3-14B cost per 1M tokens?

Llama 3.2 3B: from USD 0.0509 input and USD 0.335 output at Featherless; median of 4 standard prices USD 0.1005 / USD 0.335. Qwen3-14B: from USD 0.07 input and USD 0.28 output at SiliconFlow; median of 3 standard prices USD 0.12 / USD 0.24. Each price is read from the provider's own pricing page; the source is linked in the table.

Which providers sell both Llama 3.2 3B and Qwen3-14B?

1 provider has a standard price for both: Featherless (USD 0.1219 and USD 0.15 at 3:1). Buying each model from its own cheapest provider costs USD 0.1219 and USD 0.1225.

Which is cheaper for a coding agent or a chatbot, Llama 3.2 3B or Qwen3-14B?

For a coding agent (1,500 requests a day, 40,000 input tokens (85% cached), 2,000 output incl. reasoning): USD 121.95 a month on Llama 3.2 3B (Cloudflare Workers AI) against USD 151.20 on Qwen3-14B (SiliconFlow), a difference of USD 29.25. For a support chatbot (2,000 requests a day, 1,500 input tokens (60% cached), 300 output): USD 10.61 a month on Llama 3.2 3B (Featherless) against USD 11.34 on Qwen3-14B (SiliconFlow), a difference of USD 0.73. Each uses the cheapest standard price for that workload, cached input included. Try your own numbers.

At what input:output ratio does the cheaper of Llama 3.2 3B and Qwen3-14B change?

The cheaper model changes with the mix: below 1.6 input tokens per output token, Llama 3.2 3B is cheaper; from 1.6 to 2.9, Qwen3-14B is cheaper; above 2.9, Llama 3.2 3B is cheaper. At the site's 3:1 blend, Llama 3.2 3B is cheaper. More input per output token, as with retrieval or agents that resend long contexts, favours the lower input price; long answers favour the lower output price.

Do Llama 3.2 3B and Qwen3-14B have cached input pricing?

Llama 3.2 3B: no standard price lists a cached-input discount. Qwen3-14B: no standard price lists a cached-input discount. Cached tokens are billed at the input price where no cache price is published.

Which has the larger context window, Llama 3.2 3B or Qwen3-14B?

Qwen3-14B: up to 131,072 tokens, against 128,000 for Llama 3.2 3B, as stated on providers' pricing pages. Some providers offer less than the maximum; check the model pages before you pick the cheapest row for long prompts.

Read as MarkdownAsk ChatGPTAsk ClaudeAsk Perplexity