Liquid Inference

Liquid Inference is an auction exchange for LLM inference where providers compete to serve each request at the lowest price.

Liquid Inference is an auction exchange where providers compete on price for every AI request. Users send ordinary chat requests and set rules including model, region, speed, and minimum quality, and the cheapest provider that meets those requirements answers the request. Providers set their prices per token and change them based on supply and demand. The platform supports OpenAI Chat Completions and Anthropic Messages on one base URL so existing code works unchanged. It tests every provider regularly with standard benchmarks and publishes the results.

Liquid Inference at a glance

Best for: Software engineer
Runs on: API, Web
Pricing: Free
Released: October 2026
Category: AI coding tool
Website: liquidinference.ai

Key features

Instant auctions for every request
Routing controlled by user rules
Price limits set before the first token
Itemized billing for every request
Public price history and live data
Compatibility with OpenAI and Anthropic APIs

Use it for

Send chat requests to the cheapest available provider
Download live and past prices for models
Sell inference on models you already run
Set rules for model, region, speed, and quality

Alternatives to Liquid Inference

LiteLLM — One OpenAI-format API for 100+ LLM providers
OpenLLM — It simplifies infrastructure management and optimizes inference speed and cost.
OptiLLM — Increases LLM reasoning accuracy by up to 10x without model training.
WebLLM — Runs high-performance language models directly inside web browsers.
oMLX — It uses paged SSD KV caching to reduce agent TTFT to under 5 seconds.
AirLLM — Runs massive 70B+ parameter models on small 4GB consumer GPUs.
llm-d — Deploys production-grade LLM serving on existing hardware clusters.
Text Embeddings Inference — High-performance serving for open-source embedding models.
Resin — Builds high-performance inference engines for complex models like DeepSeek and GLM.
LightLLM — It delivers high-speed inference and serving performance for large language models.
vLLM — Maximizes throughput with PagedAttention and continuous batching.
Xinference — Supports a wide range of model types and hardware, including multi-node setups.

All Liquid Inference alternatives