Skip to content

Neuralwatt

Pricing page

43 listings across 16 models

Data policies

from the provider's published policies

Privacy policyTerms of service

On training

We do not use your API inputs or outputs to train, fine-tune, or improve any machine learning models.

Source: NeuralWatt privacy policy

On keeping your data

Inference content (prompts and completions): Processed in real-time and not retained beyond short-lived operational caching (up to 24 hours for multi-turn sessions).

Source: NeuralWatt privacy policy

More from Neuralwatt on keeping your data (1)

Analytics data (performance metrics, energy usage): Retained for up to 90 days.

Source: NeuralWatt privacy policy

We take these statements from Neuralwatt's own published policies each time we refresh the data, so they can change or disappear when those pages do. Check with the provider before you rely on any of this.

Sold by Neuralwatt

24 listings sold by Neuralwatt.

Default order. Select a score for details.

About these columns
llminfo IQ
llminfo IQ (capability index). How llminfo IQ is computed
Coding IQ
llminfo Coding IQ (coding capability, its own scale). How Coding IQ is computed
Value
Value for money. 50 means as expected for its results. Above 50, cheaper per task. How Value for money works
Yap
Tokens used per coding task vs. models with the same results. Includes agents that write code. Chattiness on maths doesn't predict it. How the Yap score works
Kimi K3.1 Pro
$2.00cache read $0.20$10.001M
GLM 5.3
$1.45cache read $0.145$4.501M
DeepSeek V4.1 Flash
$0.15cache read $0.015$0.601M
MiMo V2.6 Pro
$0.435cache read $0.0036$0.871MNot measured
GLM 5.3 Flashfp8
$0.15cache read $0.03$0.501M
Qwen 3.8 27Bfp8
$0.45cache read $0.25$3.20262KNot measured
Gemma 4 31B ITgemma-4-31b ยท fp4
$0.144cache read $0.0144$0.42262K
Neuralwatt Flashnw-flash
Not enough data yetNot enough data yet$0.15cache read $0.015$0.601MNot enough data yetNot measuredNot measured
DeepSeek V4.1 Flash (Speed)
Not enough data yetNot enough data yet$0.15cache read $0.015$0.601MNot enough data yetNot measuredNot measured
Mammoth 1
Doesnโ€™t apply to this modelDoesnโ€™t apply to this model$0.18cache read $0.018Doesnโ€™t apply to this model33KDoesnโ€™t apply to this modelDoesnโ€™t apply to this modelDoesnโ€™t apply to this model
Qwen3.6 35B
Not enough data yetNot enough data yet$0.29cache read $0.029$1.15262KNot enough data yetNot measuredNot measured
Neuralwatt Smallnw-small
Not enough data yetNot enough data yet$0.45cache read $0.25$3.20262KNot enough data yetNot measuredNot measured
Neuralwatt Largenw-large
Not enough data yetNot enough data yet$2.00cache read $0.20$10.001MNot enough data yetNot measuredNot measured
Show all 24 listings

Showing 13 of 24: one listing for each model.

Run by Neuralwatt

19 listings where Neuralwatt runs the model. Another provider sells these listings.

Default order. Select a score for details.

About these columns
llminfo IQ
llminfo IQ (capability index). How llminfo IQ is computed
Coding IQ
llminfo Coding IQ (coding capability, its own scale). How Coding IQ is computed
Value
Value for money. 50 means as expected for its results. Above 50, cheaper per task. How Value for money works
Yap
Tokens used per coding task vs. models with the same results. Includes agents that write code. Chattiness on maths doesn't predict it. How the Yap score works
Kimi K3.1 Pro๐Ÿ‡บ๐Ÿ‡ธ US
ย NanoGPT$2.10cache read $0.21$10.50Not published by the provider
DeepSeek V4.1 Flash 1017๐Ÿ‡บ๐Ÿ‡ธ US
ย NanoGPT$0.1575cache read $0.0158$0.63Not published by the provider
GLM 5.3๐Ÿ‡บ๐Ÿ‡ธ US
ย NanoGPT$1.5225cache read $0.1523$4.725Not published by the provider
MiMo V2.6 Pro๐Ÿ‡บ๐Ÿ‡ธ US
ย NanoGPT$0.4568cache read $0.0038$0.9135Not published by the providerNot measured
GLM 5.3 Flashfp8 ยท ๐Ÿ‡บ๐Ÿ‡ธ US
ย NanoGPT$0.1575cache read $0.0315$0.525Not published by the provider
Qwen 3.8 27Bfp8 ยท ๐Ÿ‡บ๐Ÿ‡ธ US
ย NanoGPT$0.4725cache read $0.2625$3.36Not published by the providerNot measured
DeepSeek V4 Flashdeepseek/deepseek-v4-flash-latest ยท latest ยท ๐Ÿ‡บ๐Ÿ‡ธ US
ย NanoGPT$0.1575cache read $0.0158$0.63Not published by the provider
Gemma 4 31B ITfp4 ยท ๐Ÿ‡บ๐Ÿ‡ธ US
ย NanoGPT$0.1512cache read $0.0151$0.441Not published by the provider
Qwen3.6 35B A3B๐Ÿ‡บ๐Ÿ‡ธ US
ย NanoGPT$0.3045cache read $0.0305$1.2075Not published by the providerNot measured
Hermes High๐Ÿ‡บ๐Ÿ‡ธ US
ย NanoGPTNot enough data yetNot enough data yet$1.5225cache read $0.1523$4.725Not published by the providerNot enough data yetNot measuredNot measured
Hermes Low๐Ÿ‡บ๐Ÿ‡ธ US
ย NanoGPTNot enough data yetNot enough data yet$1.5225cache read $0.1523$4.725Not published by the providerNot enough data yetNot measuredNot measured
Hermes Medium๐Ÿ‡บ๐Ÿ‡ธ US
ย NanoGPTNot enough data yetNot enough data yet$1.5225cache read $0.1523$4.725Not published by the providerNot enough data yetNot measuredNot measured
Show all 19 listings

Showing 12 of 19: one listing for each model.

Data credits: Epoch AI Benchmarking Hub, CC BY 4.0 (licence)

Around the web

  • Tools and recipes for using the Neuralwatt inference API.
    Neuralwatt builds software for energy-efficient AI infrastructure. Neuralwatt returns energy consumption data with every API โ€ฆNeuralwatt builds software for energy-efficient AI infrastructure. Neuralwatt returns energy consumption data with every API response, energy efficiency ...

    github.com

    Repository
  • Neuralwatt and Opencode Go for workflows
    Neuralwatt has been a surprisingly good cheap pair with Opencode Go for agentic workflows โ€ฆ โ€ฆNeuralwatt has been a surprisingly good cheap pair with Opencode Go for agentic workflows โ€ฆ its agentic coding cli made for large codebases. ...

    reddit.com

    Discussion
  • Please correct me if you have contradicting data but
    Neuralwatt's price per token vs price for energy comparison doesn't seem to take into account โ€ฆNeuralwatt's price per token vs price for energy comparison doesn't seem to take into account the cost savings from cache hits. They offer a $5 trial credit

    news.ycombinator.com

    Discussion
  • Neuralwatt boosts AI inference throughput 33%
    The startup provides intelligent AI power management software that allows customers to dramatically reduce their โ€ฆThe startup provides intelligent AI power management software that allows customers to dramatically reduce their energy consumption and ...Read more

    crusoe.ai

  • Neuralwatt Models: IDs, Pricing & Usage
    Browse Neuralwatt model IDs, capabilities, context windows, and pricing. Access 29 Neuralwatt models through Mastra's โ€ฆBrowse Neuralwatt model IDs, capabilities, context windows, and pricing. Access 29 Neuralwatt models through Mastra's model router.

    mastra.ai

How this data works