Novita model pricing.

Prices checked

Novita sells 138 prices on 101 models, from $0.02 per million tokens, checked .

Every model Novita sells

  1. 1

    Llama 3.2 1B Instruct

    Runs it themselves

    $0.02 in · $0.02 out

    $0.02/M tokens
    View
  2. 2

    Llama 3.1 8B Instruct

    Runs it themselves

    $0.02 in · $0.05 out

    $0.03/M tokens
    View
  3. 3

    Llama 3.1 8B Instruct

    Lower precision (fp8)

    $0.02 in · $0.05 out

    $0.03/M tokens
    View
  4. 4

    Llama 3.2 3B Instruct

    Runs it themselves

    $0.03 in · $0.05 out

    $0.04/M tokens
    View
  5. 5

    Qwen3 8B

    Lower precision (fp8)

    $0.04 in · $0.14 out

    $0.06/M tokens
    View
  6. 6

    Gemma 3 12B

    Same as maker

    $0.05 in · $0.10 out

    $0.06/M tokens
    View
  7. 7

    gpt-oss-20b

    Costs more

    $0.04 in · $0.15 out

    $0.07/M tokens
    View
  8. 8

    Qwen2.5 7B Instruct

    Runs it themselves

    $0.07 in · $0.07 out

    $0.07/M tokens
    View
  9. 9

    Mistral Nemo

    Runs it themselves

    $0.04 in · $0.17 out

    $0.07/M tokens
    View
  10. 10

    Nemotron 3 Nano 30B A3B

    Runs it themselves

    $0.05 in · $0.20 out

    $0.09/M tokens
    View
  11. 11

    Nemotron 3 Nano 30B A3B

    Lower precision (fp4)

    $0.05 in · $0.20 out

    $0.09/M tokens
    View
  12. 12

    Ling 3.0 Flash

    Costs more

    $0.06 in · $0.18 out

    $0.09/M tokens
    View
  13. 13

    Ling 3.0 Flash

    Costs more

    $0.06 in · $0.18 out

    $0.09/M tokens
    View
  14. 14

    MythoMax 13B

    Costs more

    $0.09 in · $0.09 out

    $0.09/M tokens
    View
  15. 15

    gpt-oss-120b

    Runs it themselves

    $0.05 in · $0.25 out

    $0.10/M tokens
    View
  16. 16

    gpt-oss-120b

    Lower precision (fp4)

    $0.05 in · $0.25 out

    $0.10/M tokens
    View
  17. 17

    $0.07 in · $0.22 out

    $0.11/M tokens
    View
  18. 18

    $0.07 in · $0.27 out

    $0.12/M tokens
    View
  19. 19

    GLM-5.3-Flash

    Lower precision (fp8)

    $0.08 in · $0.28 out

    $0.13/M tokens
    View
  20. 20

    Gemma 3 27B

    Costs more

    $0.12 in · $0.20 out

    $0.14/M tokens
    View
  21. 21

    Gemma 3 27B

    Costs more

    $0.12 in · $0.20 out

    $0.14/M tokens
    View
  22. 22

    GLM-4.7-Flash

    Costs more

    $0.07 in · $0.40 out

    $0.15/M tokens
    View
  23. 23

    GLM-4.7-Flash

    Costs more

    $0.07 in · $0.40 out

    $0.15/M tokens
    View
  24. 24

    MiMo-V2.6-Flash

    Same as maker

    $0.14 in · $0.28 out

    $0.18/M tokens
    View
  25. 25

    MiMo-V2.6-Flash

    Same as maker

    $0.14 in · $0.28 out

    $0.18/M tokens
    View

Not ranked

Below the maker's own price with nothing published to account for it, or from a catalog we could not read reliably. Why

  1. Qwen3 32B

    Not ranked

    $0.10 in · $0.45 out

    $0.19/M tokens
    View
  2. Qwen3 235B-A22B

    Not ranked

    $0.09 in · $0.58 out

    $0.21/M tokens
    View
  3. Qwen-MT Plus

    Not ranked

    $0.25 in · $0.75 out

    $0.38/M tokens
    View
  4. Qwen3.7 Max

    Not ranked

    $1.25 in · $3.75 out

    $1.88/M tokens
    View

Where to start

Kind

Host

based in US

Free tier

No

every price is paid

What Novita charges per million tokens embedded

Priced on input alone, since an embedding model writes no output, so never held beside a chat model's blended price. Every embedding model

Embedding models, per 1M input tokens

  1. 1

    BGE M3

    Runs it themselves

    Beijing Academy of Artificial Intelligence

    $0.01/M tokens
  2. 2

    Alibaba Cloud

    $0.07/M tokens
  3. 3

    Qwen3 Embedding 8B

    No maker price

    Alibaba Cloud

    $0.07/M tokens

Frequently asked

What is Novita?
Serves open-weight models on its own GPUs; publishes list and promotional prices side by side. It sells 101 of the models we track, and trades as Novita AI, based in US.
Is Novita cheaper than buying from the maker?
53 of its 138 prices are below the maker's own price, and 39 of those come with a published reason, so they can lead a ranking. 4 are listed and not ranked, because nothing accounts for the gap.
Does Novita train on what you send it?
Novita does not say whether it trains on what you send it. It does not say how long it keeps requests.
How current are these Novita prices?
They come from its own public catalog, read every 3 days; the reading on this page is from 31 Oct 2026. Prices exclude tax.

Models it sells

Other hosts

Where to go next