modelgrep

Cheapest OpenAI Models

Quick answer · Updated June 2026

The cheapest OpenAI model is gpt-oss-20b at $0.029 per million input tokens. gpt-oss-120b ($0.039) and GPT-5 Nano ($0.050) round out the top three.

$0.029Input /M
24.5Intelligence
391 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. 1O
    gpt-oss-20b
    ReasoningToolsJSON24.5 intel · 391 t/s · 247ms ttft
    $0.029
    Input /M
  2. 2O
    gpt-oss-120b
    ReasoningToolsJSON33.3 intel · 547 t/s · 177ms ttft
    $0.039
    Input /M
  3. 3O
    gpt-5-nano
    ReasoningToolsJSON+125.9 intel · 400K ctx
    $0.050
    Input /M
  4. 4O
    gpt-oss-safeguard-20b
    ReasoningToolsJSON570 t/s · 229ms ttft · 131K ctx
    $0.075
    Input /M
  5. 5O
    gpt-4.1-nano
    ToolsJSONVision13.0 intel · 91 t/s · 649ms ttft
    $0.100
    Input /M
  6. 6O
    gpt-4o-mini-search-preview
    JSON46 t/s · 2.4s ttft · 128K ctx
    $0.150
    Input /M
  7. 7O
    gpt-4o-mini-2024-07-18
    ToolsJSONVision51 t/s · 528ms ttft · 128K ctx
    $0.150
    Input /M
  8. 8O
    gpt-4o-mini
    ToolsJSONVision41 t/s · 520ms ttft · 128K ctx
    $0.150
    Input /M
  9. 9O
    gpt-5.4-nano
    ReasoningToolsJSON+144.0 intel · 60 t/s · 552ms ttft
    $0.200
    Input /M
  10. 10O
    gpt-5.1-codex-mini
    ReasoningToolsJSON+138.6 intel · 127 t/s · 1.7s ttft
    $0.250
    Input /M
  11. 11O
    gpt-5-mini
    ReasoningToolsJSON+141.2 intel · 70 t/s · 729ms ttft
    $0.250
    Input /M
  12. 12O
    gpt-4.1-mini
    ToolsJSONVision22.9 intel · 45 t/s · 689ms ttft
    $0.400
    Input /M
  13. 13O
    gpt-3.5-turbo
    ToolsJSON50 t/s · 413ms ttft · 16K ctx
    $0.500
    Input /M
  14. 14O
    gpt-audio-mini
    ToolsJSONAudio60 t/s · 459ms ttft · 128K ctx
    $0.600
    Input /M
  15. 15O
    gpt-5.4-mini
    ReasoningToolsJSON+123.3 intel · 83 t/s · 728ms ttft
    $0.750
    Input /M
  16. 16O
    gpt-3.5-turbo-0613
    ToolsJSON46 t/s · 794ms ttft · 4K ctx
    $1.00
    Input /M
  17. 17O
    o4-mini-high
    ReasoningToolsJSON+1124 t/s · 4.3s ttft · 200K ctx
    $1.10
    Input /M
  18. 18O
    o4-mini
    ReasoningToolsJSON+133.1 intel · 122 t/s · 4.5s ttft
    $1.10
    Input /M
  19. 19O
    o3-mini-high
    ReasoningToolsJSON25.2 intel · 200K ctx
    $1.10
    Input /M
  20. 20O
    o3-mini
    ReasoningToolsJSON144 t/s · 2.2s ttft · 200K ctx
    $1.10
    Input /M
  21. 21O
    gpt-5.1-codex-max
    ReasoningToolsJSON+167 t/s · 1.6s ttft · 400K ctx
    $1.25
    Input /M
  22. 22O
    gpt-5.1
    ReasoningToolsJSON+147.7 intel · 67 t/s · 1.0s ttft
    $1.25
    Input /M
  23. 23O
    gpt-5.1-chat
    ToolsJSONVision61 t/s · 1.1s ttft · 128K ctx
    $1.25
    Input /M
  24. 24O
    gpt-5.1-codex
    ReasoningToolsJSON+143.1 intel · 79 t/s · 2.8s ttft
    $1.25
    Input /M
  25. 25O
    gpt-5-codex
    ReasoningToolsJSON+144.6 intel · 43 t/s · 3.6s ttft
    $1.25
    Input /M

Frequently asked

What is the cheapest OpenAI model?

The cheapest OpenAI model is gpt-oss-20b at $0.029 per million input tokens. gpt-oss-120b ($0.039) and GPT-5 Nano ($0.050) round out the top three.

What's a good alternative to gpt-oss-20b?

gpt-oss-120b ($0.039) is the closest alternative on this metric, followed by GPT-5 Nano ($0.050). See the full ranking above for the tradeoffs.

How many OpenAI models are there?

modelgrep tracks 62 OpenAI models with live benchmarks, speed, latency and per-provider pricing, led on intelligence by GPT-5.4. 25 of them qualify for this ranking.

More OpenAI rankings

All rankings