Qwen3 32B vs GPT-5 Nano API pricing
Updated
Qwen3 32B: USD 0.08 input, USD 0.28 output per 1M tokens (DeepInfra). GPT-5 Nano: USD 0.05 input, USD 0.40 output per 1M tokens (OpenAI and 2 others). At 3 input tokens per output token, GPT-5 Nano costs 5.8% more than Qwen3 32B (USD 0.1375 vs USD 0.13 per 1M).
How do Qwen3 32B and GPT-5 Nano differ on price?
- Cheapest input price: GPT-5 Nano USD 0.05 against USD 0.08 per 1M tokens, so Qwen3 32B costs 1.6× as much (60% more).
- Cheapest output price: Qwen3 32B USD 0.28 against USD 0.40 per 1M tokens, so GPT-5 Nano costs 1.43× as much (43% more).
- Among the 201 indexed models, Qwen3 32B ranks 32nd cheapest and GPT-5 Nano 33rd on the 3:1 blend.
- The cheaper model changes at 4 input tokens per output token.
- Both list cached-input discounts: Qwen3 32B at 1 standard price of 5, from USD 0.0816 (Featherless, 20% below that provider's input price); GPT-5 Nano at 3 standard prices of 3, from USD 0.005 (OpenAI, 90% below that provider's input price).
- Only Qwen3 32B has a stated context window (131,072 tokens); no provider of GPT-5 Nano states one on its pricing page.
- Qwen3 32B has 5 providers with a standard price, 2 more than GPT-5 Nano; no provider sells both, so using both means two accounts.
- Shopping around matters more for Qwen3 32B: its dearest standard price is 2× its cheapest, against 1× for GPT-5 Nano.
- OpenAI's own API is one of 3 providers tied for the lowest standard price we found for GPT-5 Nano, USD 0.1375 at 3:1.
- Qwen3 32B has a batch price at 1 provider, from USD 0.1312 at 3:1 (Amazon Bedrock); GPT-5 Nano has a batch price at 1 provider, from USD 0.06875 at 3:1 (OpenAI).
- For the one-developer coding agent preset, the gap is USD 2,954 a year in GPT-5 Nano's favour (Qwen3 32B USD 303.30 a month, GPT-5 Nano USD 57.15).
- Per character of input, the gap depends on the language: Qwen3 32B costs 4.6× as much as GPT-5 Nano on Hindi text but 1.33× as much as GPT-5 Nano on Chinese, because the two tokenizers split those languages differently.
Qwen3 32B and GPT-5 Nano 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.
| Price | Qwen3 32B | GPT-5 Nano |
|---|---|---|
| Model page | Qwen3 32B FP8 Qwen · 5 providers | GPT-5 Nano OpenAI · 3 providers |
| Cheapest standard price, blended 3:1 | USD 0.13 lower USD 0.08 in / USD 0.28 out at DeepInfra DeepInfra page checked 2026-11-14 | USD 0.1375 USD 0.05 in / USD 0.40 out at OpenAI and 2 others OpenAI page checked 2026-11-14 |
| Cheapest input | USD 0.08 DeepInfra source | USD 0.05 lower OpenAI source |
| Cheapest cached input | USD 0.0816 Featherless source | USD 0.005 lower OpenAI source |
| Cheapest output | USD 0.28 lower DeepInfra source | USD 0.40 OpenAI source |
| Maker's own API | Qwen publishes no paid standard price on its own API that we found. | USD 0.05 / USD 0.40 OpenAI source |
| Median standard price | USD 0.102 / USD 0.57 input / output, 5 prices | USD 0.05 / USD 0.40 input / output, 3 prices |
| Context window | 131,072 tokens, largest stated; from 32,768 at some providers | No provider states one on its pricing page. |
| Providers with a standard price | 5 | 3 |
- Lower marks the lower price on a line; tie, equal prices. Other tiers and free rows are on the model pages.
Report a wrong price or number (email to info@broadnet.ai, with this page's link filled in).
What do Qwen3 32B and GPT-5 Nano 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).
| Workload | Qwen3 32B | GPT-5 Nano | Difference |
|---|---|---|---|
| Support chatbot 2,000 requests a day, 1,500 input tokens (60% cached), 300 output | USD 12.24 DeepInfra | USD 9.27 OpenAI | USD 2.97 GPT-5 Nano lower |
| RAG over documents 1,000 requests a day, 8,000 input tokens (30% cached), 500 output | USD 23.40 DeepInfra | USD 14.76 OpenAI | USD 8.64 GPT-5 Nano lower |
| Coding agent (one developer) 1,500 requests a day, 40,000 input tokens (85% cached), 2,000 output incl. reasoning | USD 303.30 SiliconFlow | USD 57.15 OpenAI | USD 246.15 GPT-5 Nano lower |
GPT-5 Nano is cheaper on all three workloads. The widest gap is on Coding agent (one developer): the dearer model costs 5.3× as much as the cheaper one.
Does the cheaper model change with your input:output ratio?
The cheaper model changes with the mix: below 4 input tokens per output token, Qwen3 32B is cheaper; above 4, GPT-5 Nano is cheaper. At the site's 3:1 blend, Qwen3 32B 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.
| Input : output tokens | Qwen3 32B | GPT-5 Nano |
|---|---|---|
| 1 : 3 (long answers) | USD 0.23 lower DeepInfra | USD 0.3125 OpenAI |
| 1 : 1 | USD 0.18 lower DeepInfra | USD 0.225 OpenAI |
| 3 : 1 (the site’s blend) | USD 0.13 lower DeepInfra | USD 0.1375 OpenAI |
| 10 : 1 (retrieval, long prompts) | USD 0.09818 DeepInfra | USD 0.08182 lower OpenAI |
| 30 : 1 (agents, big contexts) | USD 0.08645 DeepInfra | USD 0.06129 lower OpenAI |
USD per 1M tokens at the cheapest standard provider for each mix, before cache discounts. Thinking tokens bill as output.
What do Qwen3 32B and GPT-5 Nano cost per character in other languages?
Cost to send 1M characters at each model's cheapest non-promotional standard input price, with the Qwen3 32B tokenizer for Qwen3 32B and the GPT (o200k) tokenizer for GPT-5 Nano:
| Content | Qwen3 32B | GPT-5 Nano |
|---|---|---|
| English | USD 0.018 DeepInfra | USD 0.011 lower Microsoft Azure |
| Arabic MSA | USD 0.031 DeepInfra | USD 0.016 lower Microsoft Azure |
| Egyptian Arabic | USD 0.032 DeepInfra | USD 0.017 lower Microsoft Azure |
| Hindi | USD 0.074 DeepInfra | USD 0.016 lower Microsoft Azure |
| Chinese | USD 0.051 DeepInfra | USD 0.039 lower Microsoft Azure |
| Spanish | USD 0.021 DeepInfra | USD 0.011 lower Microsoft Azure |
| Code | USD 0.021 DeepInfra | USD 0.013 lower Microsoft Azure |
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
- Qwen3 32B FP8 API pricing: all 5 providers and other tiers.
- GPT-5 Nano API pricing: all 3 providers and other tiers.
- Cost calculator with Qwen3 32B and GPT-5 Nano preloaded
- Qwen3 token counter
- GPT token counter
- Prompt caching calculator
More comparisons: Gemma 4 26B A4B vs Qwen3 32B · Qwen3.5 27B vs Qwen3 32B · Qwen3 32B vs Qwen3 30B A3B · Qwen3 235B A22B vs Qwen3 32B · Gemma 4 26B A4B vs GPT-5 Nano · GPT-5.4 Nano vs GPT-5 Nano · GPT-5 Mini vs GPT-5 Nano · GPT-5 Nano vs GPT-5 · every Qwen3 32B comparison · every GPT-5 Nano comparison · all 662 price comparisons
Questions
Is Qwen3 32B cheaper than GPT-5 Nano?
Yes. On a blend of 3 input tokens to 1 output token, Qwen3 32B costs USD 0.13 per 1M tokens and GPT-5 Nano USD 0.1375, at each model's cheapest standard price. For input alone, GPT-5 Nano is cheaper (USD 0.05 against USD 0.08 per 1M); for output alone, Qwen3 32B is (USD 0.28 against USD 0.40). This compares price only, not quality.
What do Qwen3 32B and GPT-5 Nano cost per 1M tokens?
Qwen3 32B: from USD 0.08 input and USD 0.28 output at DeepInfra; median of 5 standard prices USD 0.102 / USD 0.57. GPT-5 Nano: from USD 0.05 input and USD 0.40 output at OpenAI and 2 others; OpenAI's own API charges USD 0.05 / USD 0.40; median of 3 standard prices USD 0.05 / USD 0.40. Each price is read from the provider's own pricing page; the source is linked in the table.
Which providers sell both Qwen3 32B and GPT-5 Nano?
None of the providers we track sells both at a standard price. Qwen3 32B is cheapest at DeepInfra and GPT-5 Nano at OpenAI and 2 others, so running both means two accounts.
Which is cheaper for a coding agent or a chatbot, Qwen3 32B or GPT-5 Nano?
For a coding agent (1,500 requests a day, 40,000 input tokens (85% cached), 2,000 output incl. reasoning): USD 303.30 a month on Qwen3 32B (SiliconFlow) against USD 57.15 on GPT-5 Nano (OpenAI), a difference of USD 246.15. For a support chatbot (2,000 requests a day, 1,500 input tokens (60% cached), 300 output): USD 12.24 a month on Qwen3 32B (DeepInfra) against USD 9.27 on GPT-5 Nano (OpenAI), a difference of USD 2.97. 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 Qwen3 32B and GPT-5 Nano change?
The cheaper model changes with the mix: below 4 input tokens per output token, Qwen3 32B is cheaper; above 4, GPT-5 Nano is cheaper. At the site's 3:1 blend, Qwen3 32B 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 Qwen3 32B and GPT-5 Nano have cached input pricing?
Qwen3 32B from USD 0.0816 per 1M cached input tokens (Featherless; 1 of 5 standard prices list a discount). GPT-5 Nano from USD 0.005 per 1M cached input tokens (OpenAI; 3 of 3 standard prices list a discount). Cached tokens are billed at the input price where no cache price is published.
Read as MarkdownAsk ChatGPTAsk ClaudeAsk Perplexity
Sources