Gemma 3 27B API pricing
Google · Budget price tier · Gemma · released 2025-03. Prices in USD per 1 million tokens. Data snapshot: 2026-11-12.
A support ticket workload of 100K requests a month costs $29.50 on Gemma 3 27B; the next cheaper model in the Budget price tier for this workload, GLM 4.7 Flash, does it for $29.07. The price we quote for Gemma 3 27B has changed 3 times since 2026-08-15, most recently on 2026-11-12 when the quoted host changed (-20.6%); it is now the 57th cheapest of 160 models in the Budget price tier. Buying from DeepInfra rather than at the price we quote lowers the token price by about 49%, for a quantized fp8 build. It is rated a stable release but does not meet all four criteria of the FinOps Friendly badge.
- Input: $0.08 per 1M tokens
- Output: $0.35 per 1M tokens
- Batch API (in/out): Not published
- Prompt-cache read: $0.04 per 1M tokens
- Context window: 131K tokens
- Arena ELO: 1358 as of 2026-11-08
Available on
AWS Bedrock regions: ap-northeast-1, eu-west-1, us-east-1, us-east-2, us-west-2
Market position
On blended price ($0.269 per 1M tokens at 30% input, 70% output) it is cheaper than 65% of models in the Budget price tier: 57th cheapest of 160. Across every price tier it is cheaper than 81% of priced text models in the catalogue: 57th cheapest of 303. Its Arena ELO of 1358 ranks 121st of the 160 models that carry one, above 25% of them. With an Arena score of 1358, Gemma 3 27B sits below OptimNow's lowest working quality threshold, Standard (Arena 1420+), so the quality-bar method would not shortlist it for any of the 8 use cases the table prices. Each threshold is a distance below the best ranked model, which scores 1530 on Arena scores as of 8 Nov 2026.
Licence, size and capabilities
Gemma 3 27B is open weights, which is not the same as open source: the weights are published, but under Gemma, a licence that restricts some uses. Read it before self-hosting commercially. Its published size is 27B parameters. Listed capabilities: Text, Vision, Agents. Filling its 131K context window once costs $0.0105 in input tokens at list price.
Pricing by host
A host is a company that runs the model and sells it by the token: the model maker, a cloud platform like Amazon Bedrock, Azure or Vertex, or an independent host. One model, several hosts, and often several prices. This model is sold by 4 different hosts. The price we quote elsewhere on this page is marked. We quote one host's rate so that models stay comparable; it is not a recommendation to buy there.
Cheapest host: DeepInfra at $0.08 input / $0.16 output per 1M tokens, 49% below the price we quote on the blended price (30% input, 70% output). It serves a quantized fp8 build, which is not the same product as the full-precision model. A support ticket workload at 100K requests a month costs $20.00 there against $29.50 at the price we quote, a difference of $9.50 a month. The dearest, Parasail, charges 2.5x the cheapest.
| Host | Input $/1M | Output $/1M | Precision |
|---|---|---|---|
| DeepInfra | $0.08 | $0.16 | fp8 |
| Nebius | $0.10 | $0.30 | fp8 |
| NextBit (price we quote) | $0.08 | $0.35 | fp8 |
| Parasail | $0.08 | $0.45 | fp8 |
List prices as published on 2026-11-12, via OpenRouter. Negotiated rates, committed-use discounts and provisioned capacity are not shown.
Monthly cost at 100K requests, by business use case
From the after-discounts figure (prompt caching on the cached input share, batch API for async workloads, only where Google publishes those rates), which assumes you implement both, to the list price. One figure means no discount is published.
| Use case | Tokens (in/out) | Monthly (after discounts – list) |
|---|---|---|
| Support Ticket | 1,500 / 500 | $26 – $30 |
| Knowledge Q&A | 2,000 / 800 | $38 – $44 |
| Meeting Summary | 10,000 / 1,200 | $118 – $122 |
| Marketing Content | 2,500 / 1,800 | $81 – $83 |
| Coding Task | 3,000 / 2,000 | $88 – $94 |
| Invoice Processing | 1,500 / 600 | $31 – $33 |
| Call Summary | 2,000 / 700 | $40 – $41 |
| Agent Workflow | 6,000 / 3,000 | $136 – $153 |
Price history
Gemma 3 27B has been in our daily price archive since tracking began on 2026-08-15, when it listed at $0.08 input / $0.45 output per 1M tokens. The price we quote has changed 3 times since then, every time through a change of quoted host rather than a repricing. The latest change, on 2026-11-12, was -20.6% on the blended price (30% input, 70% output), to $0.08 input / $0.35 output (quoted host changed, not a repricing). Net of all changes, the blended price is -20.6% against 2026-08-15.
| Date | Old price (in / out) | New price (in / out) | Blended change |
|---|---|---|---|
| 2026-11-12 | $0.08 / $0.45 | $0.08 / $0.35 | -20.6% (quoted host changed, not a repricing) |
| 2026-11-09 | $0.08 / $0.35 | $0.08 / $0.45 | +26.0% (quoted host changed, not a repricing) |
| 2026-10-18 | $0.08 / $0.45 | $0.08 / $0.35 | -20.6% (quoted host changed, not a repricing) |
Alternatives to Gemma 3 27B
None of the 4 comparable models below costs less than Gemma 3 27B on the blended price.
- Next cheaper model in the Budget price tier: Llama 3.2 3B Instruct (Meta) at $0.246 blended, 9% less.
- Next more expensive: DeepSeek V4.1 Flash (DeepSeek) at $0.2866 blended, 7% more.
The nearest models in the Budget price tier on Arena ELO, each with its blended price and score difference against Gemma 3 27B.
- MiniMax M2.5 (MiniMax): $0.27 in / $1.08 out per 1M, Arena ELO 1359 (211% more expensive, 1 ELO point higher)
- Qwen3 Coder 480B A35B (Alibaba): $0.30 in / $1.00 out per 1M, Arena ELO 1357 (194% more expensive, 1 ELO point lower)
- gpt-oss-120b (OpenAI): $0.15 in / $0.60 out per 1M, Arena ELO 1365 (73% more expensive, 7 ELO points higher)
- GLM 4.7 Flash (Zhipu): $0.0605 in / $0.40 out per 1M, Arena ELO 1351 (11% more expensive, 7 ELO points lower)
- Cheapest within 15 Arena points: Nemotron 3 Nano 30B A3B (Nvidia), 31% cheaper, 10 ELO points lower, out of 10 models in that band.
- Highest Arena ELO within 20% of this price: DeepSeek V4.1 Flash (DeepSeek), 7% more expensive, 105 ELO points higher, out of 13 models in that price band.
Other models from Google
The 8 nearest to Gemma 3 27B in price, of 27 other Google models in the catalogue. Prices per 1M tokens, input then output.
- Gemma 3 4B: $0.05 / $0.10
- Gemma 4 26B A4B: $0.08 / $0.10
- Gemma 3 12B: $0.05 / $0.15
- Gemini 2.5 Flash Lite: $0.10 / $0.40
- Gemma 4 31B: $0.20 / $0.40
- Gemma 2 27B: $0.65 / $0.65
- Gemini 3.1 Flash Lite: $0.25 / $1.50
- Nano Banana 2 Lite (Gemini 3.1 Flash Lite Image): $0.25 / $1.50
All 28 Google models and prices
Common questions about Gemma 3 27B pricing
- How much does Gemma 3 27B cost?
- Gemma 3 27B costs $0.08 per 1M input tokens and $0.35 per 1M output tokens at NextBit, the host whose rate we quote. NextBit serves a quantized fp8 build, which is not the same product as the full-precision model. Prices as listed on 2026-11-12.
- Which host is cheapest for Gemma 3 27B?
- Cheapest host: DeepInfra at $0.08 input / $0.16 output per 1M tokens, 49% below the price we quote on the blended price (30% input, 70% output). It serves a quantized fp8 build, which is not the same product as the full-precision model. A support ticket workload at 100K requests a month costs $20.00 there against $29.50 at the price we quote, a difference of $9.50 a month. The dearest, Parasail, charges 2.5x the cheapest. Host rates as listed on 2026-11-12.
- Does Gemma 3 27B offer batch or cached pricing?
- Yes, for prompt caching only. Cache reads cost $0.04 per 1M tokens, as listed on 2026-11-12. No batch pricing is published.
- What would Gemma 3 27B cost per month?
- With the Support Ticket profile (1,500 input and 500 output tokens per request), 100,000 requests a month cost about $30 on Gemma 3 27B at the list price quoted on 2026-11-12.
- Can Gemma 3 27B be self-hosted?
- Yes, within the terms of its licence. Gemma 3 27B is open weights, which is not the same as open source: the weights are published, but under Gemma, a licence that restricts some uses. Read it before self-hosting commercially.
- Has the price of Gemma 3 27B changed?
- Gemma 3 27B has been in our daily price archive since tracking began on 2026-08-15, when it listed at $0.08 input / $0.45 output per 1M tokens. The price we quote has changed 3 times since then, every time through a change of quoted host rather than a repricing. The latest change, on 2026-11-12, was -20.6% on the blended price (30% input, 70% output), to $0.08 input / $0.35 output (quoted host changed, not a repricing). Net of all changes, the blended price is -20.6% against 2026-08-15.
Its weights are published, which changes what you may run yourself: the cost gap between open weights and proprietary APIs.
These are list prices. Your real cost also depends on caching and batching, negotiated discounts, marketplace agreements, and the contracts you already have. OptimNow audits exactly that.
Build the business case · Get a FinOps review
LLM pricing · Models by maker · Cloud compute pricing · Compare models · Price barometer · FinOps guides · API and MCP server