modelgrep

Cheapest Google Models

Quick answer · Updated June 2026

The cheapest Google model is Gemma 3 4B at $0.050 per million input tokens. Gemma 3 12B ($0.050) and Gemma 4 26B A4B ($0.060) round out the top three.

$0.050Input /M
6.3Intelligence
20 t/sSpeed
131KContext

AI models ranked by input token price. The most affordable large language model APIs, from budget open-weight models to discounted frontier models.

  1. 1G
    gemma-3-4b-it
    JSONVision6.3 intel · 20 t/s · 566ms ttft
    $0.050
    Input /M
  2. 2G
    gemma-3-12b-it
    ToolsJSONVision8.8 intel · 37 t/s · 501ms ttft
    $0.050
    Input /M
  3. 3G
    gemma-4-26b-a4b-it
    ReasoningToolsJSON+131.2 intel · 68 t/s · 356ms ttft
    $0.060
    Input /M
  4. 4G
    gemma-3n-e4b-it
    35 t/s · 268ms ttft · 33K ctx
    $0.060
    Input /M
  5. 5G
    gemma-3-27b-it
    ToolsJSONVision10.3 intel · 49 t/s · 434ms ttft
    $0.080
    Input /M
  6. 6G
    gemini-2.5-flash-lite-preview-09-2025
    ReasoningToolsJSON+219.4 intel · 217 t/s · 385ms ttft
    $0.100
    Input /M
  7. 7G
    gemini-2.5-flash-lite
    ReasoningToolsJSON+217.6 intel · 118 t/s · 369ms ttft
    $0.100
    Input /M
  8. 8G
    gemma-4-31b-it
    ReasoningToolsJSON+139.2 intel · 55 t/s · 309ms ttft
    $0.120
    Input /M
  9. 9G
    gemini-3.1-flash-lite
    ReasoningToolsJSON+2116 t/s · 611ms ttft · 1.0M ctx
    $0.250
    Input /M
  10. 10G
    gemini-3.1-flash-lite-preview
    ReasoningToolsJSON+233.5 intel · 96 t/s · 624ms ttft
    $0.250
    Input /M
  11. 11G
    gemini-2.5-flash-image
    JSONVisionImage out224 t/s · 894ms ttft · 33K ctx
    $0.300
    Input /M
  12. 12G
    gemini-2.5-flash
    ReasoningToolsJSON+291 t/s · 601ms ttft · 1.0M ctx
    $0.300
    Input /M
  13. 13G
    gemini-3.1-flash-image-preview
    ReasoningJSONVision+1144 t/s · 9.7s ttft · 131K ctx
    $0.500
    Input /M
  14. 14G
    gemini-3-flash-preview
    ReasoningToolsJSON+246.4 intel · 66 t/s · 1.2s ttft
    $0.500
    Input /M
  15. 15G
    gemma-2-27b-it
    JSON37 t/s · 1.2s ttft · 8K ctx
    $0.650
    Input /M
  16. 16G
    gemini-2.5-pro
    ReasoningToolsJSON+234.6 intel · 100 t/s · 1.5s ttft
    $1.25
    Input /M
  17. 17G
    gemini-2.5-pro-preview
    ReasoningToolsJSON+2100 t/s · 1.5s ttft · 1.0M ctx
    $1.25
    Input /M
  18. 18G
    gemini-2.5-pro-preview-05-06
    ReasoningToolsJSON+2100 t/s · 1.5s ttft · 1.0M ctx
    $1.25
    Input /M
  19. 19G
    gemini-3.5-flash
    ReasoningToolsJSON+243.3 intel · 156 t/s · 1.7s ttft
    $1.50
    Input /M
  20. 20G
    gemini-3.1-pro-preview-customtools
    ReasoningToolsJSON+287 t/s · 2.9s ttft · 1.0M ctx
    $2.00
    Input /M
  21. 21G
    gemini-3.1-pro-preview
    ReasoningToolsJSON+241.3 intel · 100 t/s · 2.9s ttft
    $2.00
    Input /M
  22. 22G
    gemini-3-pro-image-preview
    ReasoningJSONVision+179 t/s · 3.5s ttft · 66K ctx
    $2.00
    Input /M

Frequently asked

What is the cheapest Google model?

The cheapest Google model is Gemma 3 4B at $0.050 per million input tokens. Gemma 3 12B ($0.050) and Gemma 4 26B A4B ($0.060) round out the top three.

What's a good alternative to Gemma 3 4B?

Gemma 3 12B ($0.050) is the closest alternative on this metric, followed by Gemma 4 26B A4B ($0.060). See the full ranking above for the tradeoffs.

How many Google models are there?

modelgrep tracks 26 Google models with live benchmarks, speed, latency and per-provider pricing, led on intelligence by Gemini 3 Flash Preview. 22 of them qualify for this ranking.

More Google rankings

All rankings