Skip to content
Pareto Inference

We tune your custom or open-⁠source model inference to your workload and quality standards: faster responses and higher limits for your users, better margins for you.

Available on

VercelOpenCodeCommand CodeFreebuffMergeVenice

Supported by

Recent news

Pareto raises $4M in seed funding to expand its GPU fleet.

Service

We benchmark your traffic, tune the stack around it and run it on dedicated capacity, with our engineers a Slack message away.

Bring your model
Custom or fine-tuned weights, served privately on our GPUs.
No GPU contract
A monthly spend minimum with the same rate on every token above it.
Dedicated capacity
Your own capacity, backed by an uptime SLA.
A direct line
A shared Slack channel and a dashboard for each model.

Scale

Up 309x from about 10B a day in early October. Capacity that ramps as fast as you grow.

Performance

Each one tuned to its own traffic, with no two clients running the same stack.

Last hour · P50Checking sample time

GLM 5.3 Flash

Uptime
100.0%
TTFT
0.75 s
TPS
115.9
Cached input
96.1%

DeepSeek V4.1 Flash

Uptime
100.0%
TTFT
1.22 s
TPS
118.0
Cached input
97.1%

Kimi K3.1 Pro

Uptime
100.0%
TTFT
0.85 s
TPS
124.6
Cached input
95.5%
Uptime history · all models12h

Pricing

Self-serve rates on our hosted open models, with the savings passed on to you and your users.

Self-serve · USD per 1M tokens
ModelHoursInputOutputCached input
GLM 5.3 Flash60% under standard pricingStandard price $0.15$0.06Standard price $0.50$0.20Standard price $0.03$0.012
DeepSeek V4.1 Flash30% under standard pricingOff-peakStandard price $0.15$0.105Standard price $0.60$0.42Standard price $0.003$0.0021
PeakStandard price $0.30$0.21Standard price $1.20$0.84Standard price $0.006$0.0042
Kimi K3.1 Pro30% under standard pricingStandard price $2.00$1.40Standard price $10.00$7.00Standard price $0.20$0.14

Quickstart

An OpenAI-compatible API: plug in your coding agent, router or app in minutes. Pay per token from one prepaid balance.

Direct API
Coding agents
and more
Routers
and more
Chat apps

Install the OpenAI Python SDK and set PARETO_API_KEY in your environment.

example.pyPython
import os
from openai import OpenAI

client = OpenAI(
    base_url="https://api.paretoinference.com/v1",
    api_key=os.environ["PARETO_API_KEY"],
)

response = client.chat.completions.create(
    model="z-ai/glm-5.3-flash",
    messages=[{"role": "user", "content": "Hello, Pareto."}],
    stream=True,
)
for chunk in response:
    if chunk.choices and chunk.choices[0].delta.content:
        print(chunk.choices[0].delta.content, end="")
Get started now