Inference.net

What it does

Inference.net is an inference platform for teams running LLMs in production. It takes in LLM requests and traces from your application or agents, routes model calls through its Gateway, and produces request metrics, cost and usage views, quality evaluations, and deployable custom or open-source model endpoints.

Tracing captures LLM calls, tool calls, and agent steps, and includes Halo, an open-source agent-loop optimizer that reads traces and suggests fixes to prompts, tools, and the harness. Gateway captures requests, tracks usage and cost, builds datasets, runs model comparisons, and supports fine-tuning and deploying task-specific models on dedicated GPUs. The two products can be used separately or together.

This suits engineering teams shipping AI features who want visibility into production LLM behavior or want to move workloads to open-source models. Pay as you go starts at $0 plus usage and Growth is $250 per month, while Enterprise requires contacting sales. Request rate limits and data retention vary by plan.

What it can do

• Capture traces of LLM calls, tool calls, and agent steps
• Route LLM traffic through Inference Gateway to record requests and usage
• Monitor latency, cost, error rates, and payload sizes
• Build datasets from production traces and run evaluations
• Compare candidate models on live traffic with AutoEvals
• Fine-tune and deploy custom models on dedicated GPUs

Before you choose

• Pay as you go includes a 1M Gateway request allowance and 14-day data retention; Growth raises limits and retention
• Dedicated GPU deployments are priced per hour by instance type and are separate from plan fees
• Enterprise custom contracts and dedicated infrastructure require contacting sales

Listed under

• Coding
• DevOps and observability
• Developers
• Fine-tuning
• Inference
• LLM infrastructure
• Model deployment
• Observability

Get started

Sign up at the register page, then install the tracing SDK.

1. Create an account at the register page
2. Install the tracing SDK from the docs and capture your first trace
3. Route traffic through Inference Gateway to record LLM calls and view metrics

Pay as you go is usage-based with 1M included Gateway requests and 1M tracing spans per month. Growth is $250 per month and includes a $50 one-time opening credit. Enterprise pricing is custom.

Source: inference.net/pricing, checked 29 Oct 2026. Prices can change; check before you buy.

Sources for this overview

Product pages, documentation, and pricing checked 29 Oct 2026. Each source supports part of the overview; availability and prices can change.

• Inference.net | Inference infrastructure for AI-native teams ↗inference.net
• Inference platform by Inference.net - Inference.net Documentation ↗docs.inference.net
• Pricing | Inference.net ↗inference.net

Company website

inference.net: product, pricing and policy pages

YC directory mirror

Batch and company name

Social accounts

Accounts the company links, and its posts on X

Human review

Checked by a person

knowledge_base

manual_review

Last checked 29 Oct 2026. Every field keeps its source and date; unknown means we couldn’t confirm it.