Gemma 4 31B vs Gemma 3 27B API pricing
Updated
Gemma 4 31B: USD 0.10 input, USD 0.34 output per 1M tokens (CoreWeave Forge). Gemma 3 27B: USD 0.08 input, USD 0.16 output per 1M tokens (DeepInfra). At 3 input tokens per output token, Gemma 4 31B costs 60% more than Gemma 3 27B (USD 0.16 vs USD 0.10 per 1M).
How do Gemma 4 31B and Gemma 3 27B differ on price?
- Cheapest input price: Gemma 3 27B USD 0.08 against USD 0.10 per 1M tokens, so Gemma 4 31B costs 1.25× as much (25% more).
- Cheapest output price: Gemma 3 27B USD 0.16 against USD 0.34 per 1M tokens, so Gemma 4 31B costs 2.1× as much.
- Among the 201 indexed models, Gemma 4 31B ranks 49th cheapest and Gemma 3 27B 21st on the 3:1 blend.
- No input:output mix changes the answer: Gemma 3 27B is cheaper at every mix before cache discounts.
- Both list cached-input discounts: Gemma 4 31B at 6 standard prices of 14, from USD 0.06 (Parasail, 60% below that provider's input price); Gemma 3 27B at 2 standard prices of 6, from USD 0.04 (Parasail, 50% below that provider's input price).
- Gemma 4 31B states up to 262,144 tokens of context, 1.32× the 198,000 of Gemma 3 27B.
- Gemma 4 31B has 14 providers with a standard price, 8 more than Gemma 3 27B; 6 providers sell both (Venice, Featherless, Novita AI, DeepInfra and 2 more).
- On one account, Venice charges USD 0.18 for Gemma 4 31B (1.12× its cheapest price) and USD 0.14 for Gemma 3 27B (1.4× its cheapest price) at 3:1, the lowest combined price among the providers that sell both.
- Shopping around matters more for Gemma 4 31B: its dearest standard price is 5.1× its cheapest, against 2.7× for Gemma 3 27B.
- Gemma 4 31B has a batch price at 1 provider, from USD 0.09 at 3:1 (Doubleword); Gemma 3 27B has no batch price on file.
- For the one-developer coding agent preset, the gap is USD 540.00 a year in Gemma 3 27B's favour (Gemma 4 31B USD 168.30 a month, Gemma 3 27B USD 123.30).
- Per character of input, Gemma 4 31B costs 1.25× as much as Gemma 3 27B in every measured language: the two tokenizers split these texts into about the same number of tokens.
Gemma 4 31B and Gemma 3 27B 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 | Gemma 4 31B | Gemma 3 27B |
|---|---|---|
| Model page | Gemma 4 31B Google · 14 providers | Gemma 3 27B Google · 6 providers |
| Cheapest standard price, blended 3:1 | USD 0.16 USD 0.10 in / USD 0.34 out at CoreWeave Forge CoreWeave Forge page checked 2026-11-15 | USD 0.10 lower USD 0.08 in / USD 0.16 out at DeepInfra DeepInfra page checked 2026-11-14 |
| Cheapest input | USD 0.10 CoreWeave Forge source | USD 0.08 lower DeepInfra source |
| Cheapest cached input | USD 0.06 Parasail source | USD 0.04 lower Parasail source |
| Cheapest output | USD 0.34 CoreWeave Forge source | USD 0.16 lower DeepInfra source |
| Maker's own API | Google publishes no paid standard price on its own API that we found. | Google publishes no paid standard price on its own API that we found. |
| Median standard price | USD 0.14 / USD 0.40 input / output, 14 prices | USD 0.1095 / USD 0.25 input / output, 6 prices |
| Context window | 262,144 tokens, largest stated; from 32,768 at some providers | 198,000 tokens, largest stated; from 32,768 at some providers |
| Providers with a standard price | 14 | 6 |
- 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 Gemma 4 31B and Gemma 3 27B 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 | Gemma 4 31B | Gemma 3 27B | Difference |
|---|---|---|---|
| Support chatbot 2,000 requests a day, 1,500 input tokens (60% cached), 300 output | USD 15.12 CoreWeave Forge | USD 10.08 DeepInfra | USD 5.04 Gemma 3 27B lower |
| RAG over documents 1,000 requests a day, 8,000 input tokens (30% cached), 500 output | USD 29.10 CoreWeave Forge | USD 21.60 DeepInfra | USD 7.50 Gemma 3 27B lower |
| Coding agent (one developer) 1,500 requests a day, 40,000 input tokens (85% cached), 2,000 output incl. reasoning | USD 168.30 Parasail | USD 123.30 Parasail | USD 45.00 Gemma 3 27B lower |
Gemma 3 27B is cheaper on all three workloads. The widest gap is on Support chatbot: the dearer model costs 50% more than the cheaper one.
Does the cheaper model change with your input:output ratio?
Gemma 3 27B is cheaper at every mix of input and output tokens, before cache discounts: its cheapest input price and its cheapest output price are both at or below Gemma 4 31B's, so the input:output ratio never changes the answer.
| Input : output tokens | Gemma 4 31B | Gemma 3 27B |
|---|---|---|
| 1 : 3 (long answers) | USD 0.28 CoreWeave Forge | USD 0.14 lower DeepInfra |
| 1 : 1 | USD 0.22 CoreWeave Forge | USD 0.12 lower DeepInfra |
| 3 : 1 (the site’s blend) | USD 0.16 CoreWeave Forge | USD 0.10 lower DeepInfra |
| 10 : 1 (retrieval, long prompts) | USD 0.1218 CoreWeave Forge | USD 0.08727 lower DeepInfra |
| 30 : 1 (agents, big contexts) | USD 0.1077 CoreWeave Forge | USD 0.08258 lower DeepInfra |
USD per 1M tokens at the cheapest standard provider for each mix, before cache discounts. Thinking tokens bill as output.
What do Gemma 4 31B and Gemma 3 27B cost per character in other languages?
Cost to send 1M characters at each model's cheapest non-promotional standard input price, with the Gemma 4 VL 26B tokenizer for Gemma 4 31B and the Gemma 3 27B tokenizer for Gemma 3 27B:
| Content | Gemma 4 31B | Gemma 3 27B |
|---|---|---|
| English | USD 0.022 CoreWeave Forge | USD 0.017 lower DeepInfra |
| Arabic MSA | USD 0.034 CoreWeave Forge | USD 0.027 lower DeepInfra |
| Egyptian Arabic | USD 0.034 CoreWeave Forge | USD 0.027 lower DeepInfra |
| Hindi | USD 0.027 CoreWeave Forge | USD 0.022 lower DeepInfra |
| Chinese | USD 0.067 CoreWeave Forge | USD 0.054 lower DeepInfra |
| Spanish | USD 0.022 CoreWeave Forge | USD 0.018 lower DeepInfra |
| Code | USD 0.029 CoreWeave Forge | USD 0.023 lower DeepInfra |
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
- Gemma 4 31B API pricing: all 14 providers and other tiers.
- Gemma 3 27B API pricing: all 6 providers and other tiers.
- Cost calculator with Gemma 4 31B and Gemma 3 27B preloaded
- Gemma token counter
- Gemma 3 token counter
- Prompt caching calculator
More comparisons: Llama 4 Scout vs Gemma 4 31B · Gemma 4 26B A4B vs Gemma 4 31B · Gemma 4 31B vs Qwen3.8 27B · DeepSeek V4.1 Flash vs Gemma 4 31B · Mistral Small 3.2 vs Gemma 3 27B · Gemma 3 27B vs Gemma 3 12B · every Gemma 4 31B comparison · every Gemma 3 27B comparison · all 662 price comparisons
Questions
Is Gemma 4 31B cheaper than Gemma 3 27B?
No. On a blend of 3 input tokens to 1 output token, Gemma 3 27B costs USD 0.10 per 1M tokens and Gemma 4 31B USD 0.16, at each model's cheapest standard price. For input alone, Gemma 3 27B is cheaper (USD 0.08 against USD 0.10 per 1M); for output alone, Gemma 3 27B is (USD 0.16 against USD 0.34). This compares price only, not quality.
What do Gemma 4 31B and Gemma 3 27B cost per 1M tokens?
Gemma 4 31B: from USD 0.10 input and USD 0.34 output at CoreWeave Forge; median of 14 standard prices USD 0.14 / USD 0.40. Gemma 3 27B: from USD 0.08 input and USD 0.16 output at DeepInfra; median of 6 standard prices USD 0.1095 / USD 0.25. Each price is read from the provider's own pricing page; the source is linked in the table.
Which providers sell both Gemma 4 31B and Gemma 3 27B?
6 providers have a standard price for both: Venice (USD 0.18 and USD 0.14 at 3:1), Featherless (USD 0.18 and USD 0.15 at 3:1), Novita AI (USD 0.205 and USD 0.1392 at 3:1), DeepInfra (USD 0.25 and USD 0.10 at 3:1), Parasail (USD 0.2125 and USD 0.1725 at 3:1), Amazon Bedrock (USD 0.205 and USD 0.2675 at 3:1). The first is the lowest combined price. Buying each model from its own cheapest provider costs USD 0.16 and USD 0.10.
Which is cheaper for a coding agent or a chatbot, Gemma 4 31B or Gemma 3 27B?
For a coding agent (1,500 requests a day, 40,000 input tokens (85% cached), 2,000 output incl. reasoning): USD 168.30 a month on Gemma 4 31B (Parasail) against USD 123.30 on Gemma 3 27B (Parasail), a difference of USD 45.00. For a support chatbot (2,000 requests a day, 1,500 input tokens (60% cached), 300 output): USD 15.12 a month on Gemma 4 31B (CoreWeave Forge) against USD 10.08 on Gemma 3 27B (DeepInfra), a difference of USD 5.04. 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 Gemma 4 31B and Gemma 3 27B change?
Gemma 3 27B is cheaper at every mix of input and output tokens, before cache discounts: its cheapest input price and its cheapest output price are both at or below Gemma 4 31B's, so the input:output ratio never changes the answer.
Do Gemma 4 31B and Gemma 3 27B have cached input pricing?
Gemma 4 31B from USD 0.06 per 1M cached input tokens (Parasail; 6 of 14 standard prices list a discount). Gemma 3 27B from USD 0.04 per 1M cached input tokens (Parasail; 2 of 6 standard prices list a discount). Cached tokens are billed at the input price where no cache price is published.
Which has the larger context window, Gemma 4 31B or Gemma 3 27B?
Gemma 4 31B: up to 262,144 tokens, against 198,000 for Gemma 3 27B, 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
Sources
- CoreWeave Forge: pricing (checked 2026-11-15)
- Parasail: pricing (checked 2026-11-14)
- DeepInfra: pricing (checked 2026-11-14)
- Venice: pricing (checked 2026-11-15)
- Featherless: pricing (checked 2026-11-15)
- Featherless: pricing (checked 2026-11-15)
- Novita AI: pricing (checked 2026-11-14)
- DeepInfra: pricing (checked 2026-11-14)
- Amazon Bedrock: pricing (checked 2026-11-14)
- Doubleword: pricing (checked 2026-11-16)