NeuralWatt
Hosted, OpenAI-compatible inference with energy-based pricing: pay a flat $5.00/kWh for actual GPU energy consumed (up to 95% cheaper on efficient models) instead of per-token, with real per-request energy metrics on every response. Standard per-token pricing and kWh-based monthly subscriptions ($20/$50/$100) are also available. Powered by Neuralwatt Optimize, also available for self-hosted vLLM via Neuralwatt Deploy.
At a glance: NeuralWatt has 8 models listed on Inference Hub. The cheapest text model is Gemma 4 31B at $0.144 input and $0.420 output per 1M tokens, and the most expensive is Kimi K3.1 Pro at $2.00 input and $10.00 output per 1M tokens. There is no free tier. Prices are checked against NeuralWatt's own pricing every week.
Features
Compliance
None listed
Pricing Model
per tokenDetails
https://api.neuralwatt.com/v1 Model Catalog (14)
| Model | Type | Input $/1M | Output $/1M | Context | Speed | Status |
|---|---|---|---|---|---|---|
| DeepSeek V4 Flash DeepSeek | llm | $0.140 | $0.280 | 1.0M | — | |
| DeepSeek V4.1 Flash DeepSeek | llm | $0.150 | $0.600 | 1.0M | — | |
| GLM 5.2 Z.ai | llm | $1.45 | $4.50 | 1.0M | — | |
| GLM 5.3 Z.ai | llm | $1.45 | $4.50 | 1.0M | — | |
| GLM 5.3 Flash Z.ai | llm | $0.150 | $0.500 | 1.0M | — | |
| Gemma 4 31B Google · 27B | llm | $0.144 | $0.420 | 262k | — | |
| Kimi K2.6 Moonshot AI · 1T MoE (32B active) | llm | $0.690 | $3.22 | 262k | — | |
| Kimi K2.7 Code Moonshot AI · 1T MoE (32B active) | code | $0.950 | $4.00 | 262k | — | |
| Kimi K3 Moonshot AI | llm | $3.00 | $15.00 | 1.0M | — | |
| Kimi K3.1 Pro Moonshot AI | llm | $2.00 | $10.00 | 1.0M | — | |
| MiMo V2.6 Pro Xiaomi | llm | $0.435 | $0.870 | 1.0M | — | |
| Qwen3.5 397B A17B Alibaba | llm | $0.690 | $4.14 | 262k | — | |
| Qwen3.6 35B A3B Alibaba | llm | $0.290 | $1.15 | 262k | — | |
| Qwen3.8 27B Alibaba | llm | $0.450 | $3.20 | 262k | — |
NeuralWatt guides
NeuralWatt API FAQ
Is the NeuralWatt API free?
Not as far as we track: NeuralWatt has no free API tier on record. The cheapest option we track is DeepSeek V4 Flash at $0.14 in / $0.28 out per 1M tokens.
What is the cheapest NeuralWatt model?
DeepSeek V4 Flash at $0.14 in / $0.28 out per 1M tokens, based on the 14 models we track for NeuralWatt.
How many models does NeuralWatt offer?
We track 14 NeuralWatt models: 13 LLM, 1 coding.
Is the NeuralWatt API OpenAI-compatible?
Yes. NeuralWatt exposes an OpenAI-compatible endpoint, so you can use the OpenAI SDKs by changing the base URL and API key.
Run NeuralWatt? Add a "Prices tracked on Inference Hub" badge to your site, or email hello@inferencehub.org if a price here is wrong.