RelayGPU
OpenAI-compatible inference API by OpenGPU.
RelayGPU sells inference on Liquid Inference. A buyer names one of its models as relaygpu/<model>, or names the model and pays the lowest offer that meets the buyer's rules.
Facts
- Handle
- relaygpu
- Website
- https://relaygpu.com/
- On the exchange since
- 2026-10-22
- Models
- 7
- Offers
- 7
- Regions
- global
- Precisions (declared)
- provider-managed
- Lowest price on
- 7 of 7 models
Reliability, last 3 days
- Probes answered
- 82.76%
- Hours with an answer
- 100.0%
- Median response time
- 3.4s
Checked by the exchange
0 of 7 offers match the fingerprint of the model. 0 have a finished evaluation.
The exchange sends fixed prompts to each offer and compares the answers with a baseline for the model. It also runs benchmarks against the offers and keeps the results.
Data handling
- Kept 90 days, declared by the provider
- No training, declared by the provider
Models
- GPT 5.4: declared precision provider-managed, global, $2.500 input and $15.000 output per million tokens, $2.500 cache read per million tokens, 272K tokens of context
- GPT 5.4 Mini: declared precision provider-managed, global, $0.750 input and $4.500 output per million tokens, $0.750 cache read per million tokens, 272K tokens of context
- GPT 5.5: declared precision provider-managed, global, $5.000 input and $30.000 output per million tokens, $0.500 cache read per million tokens, 272K tokens of context
- GPT 6 Astra: declared precision provider-managed, global, $20.000 input and $75.000 output per million tokens, $2.000 cache read per million tokens, 1M tokens of context
- Nemotron 3 Nano 30B a3b: declared precision provider-managed, global, $0.050 input and $0.200 output per million tokens, $0.050 cache read per million tokens, 128K tokens of context
- Qwen 3.6 Plus: declared precision provider-managed, global, $0.280 input and $1.680 output per million tokens, $0.028 cache read per million tokens, 64K tokens of context
- Qwen 3.8 Max: declared precision provider-managed, global, $1.716 input and $5.148 output per million tokens, $1.716 cache read per million tokens, 983.6K tokens of context