UNIFIED AI GATEWAY & CLI FOR DEVELOPERS

Code Anywhere.

Built forTermux

Turn your phone into an AI engineering station. Use free models or connect any provider with your Fusion Gateway key.

$curl -fsSL https://fusioncode.app/install | bash
Copy and paste into terminal to try

How do I get started?

Top up your wallet, grab an API key, and stream frontier models in under a minute.

Fusion CLI

Free Harness
  • Installs on Termux, Linux, macOS, and Windows
  • Run LLM prompts directly from your terminal
  • Tracks session history and git commits
  • Automate terminal workflows with custom scripts

Open-source CLI • 100% Free

Install CLI

AI Gateway

Flagship API
  • Drop-in OpenAI /v1/chat/completions replacement
  • Stream MiniMax-M2.7 & DeepSeek-V4-Flash with 1 key
  • Prompt caching saves up to 90% on tokens
  • Sub-2.4s global edge response latency

From $5.00 / 1M tokens • Pay per token

Connect Gateway
UNIFIED AI ARCHITECTURE

Built for the future of AI engineering

OpenAI-compatible API

Replace your OpenAI key with Fusion anywhere in your code for instant access to top models.

Smart Edge Routing

Cuts response latency and API token costs by up to 90%.

Transparent per-token billing

Pay for what you use. No hidden fees, no idle compute charges, and no monthly minimums.

Enterprise security

Control team permissions per key, audit API activity, and set strict access policies.

BYOK & multi-key management

Use your own OpenAI key or Fusion tokens. Control team access per project.

For Builders, by Builders

One-command CLI & Termux install

SDKs, APIs & CLI Tools

Native support for Hono, Bun & Cloudflare

Free Gateway Credits with our Builder Program

AI sovereignty for every enterprise

Comply with data localization laws
Avoid vendor lock-in
Build confidential AI/ML pipelines
Works with existing infrastructure
GPUAI
1
Get your API key
2
Send your first request
3
Scale without limits

Run top AI models like MiniMax-M2.7 and DeepSeek-V4-Flash with one OpenAI-compatible API key. Save up to 80% with no setup fees or vendor lock-in.

Who's Building on Fusion?

AI & ML Startups

Ship faster with direct access to MiniMax-M2.7 & DeepSeek-V4-Flash via a single OpenAI-compatible endpoint.

App Developers

Drop in a Fusion API key and stream frontier models without managing multiple provider accounts.

DeFi & Web3 Apps

Censorship-resistant API access with per-key rate limits, usage quotas, and transparent billing.

Enterprises

BYOK support, audit logs, and team-level API key management — built for production workloads.

One API key, multiple frontier models, transparent per-token pricing. Fusion Gateway is the fastest way to add production-grade LLM capabilities to any app — no infrastructure to manage, no rate limit surprises.

Loved by builders

Don't just take our word for it. Real feedback from the community.

Running Kimi K2.6 on my Android phone via Termux with Fusion CLI feels like pure magic.

Aung Myat Moe
Aung Myat Moe@theaungmyatmoe
TWEET

Unified gateway latency is insane. Shifted our entire team off individual API keys to Fusion.

Alex Rivers
Alex Rivers@arivers_dev
DM

Finally an open source agentic development environment that runs standalone without bloat.

Elena Rostova
Elena Rostova@elena_ai
TWEET

The ~4MB Rust binary starts instantly inside Termux. I build microservices on the train now.

Kenji Sato
Kenji Sato@kenji_rust
DM

Running Kimi K2.6 on my Android phone via Termux with Fusion CLI feels like pure magic.

Aung Myat Moe
Aung Myat Moe@theaungmyatmoe
TWEET

Unified gateway latency is insane. Shifted our entire team off individual API keys to Fusion.

Alex Rivers
Alex Rivers@arivers_dev
DM

Finally an open source agentic development environment that runs standalone without bloat.

Elena Rostova
Elena Rostova@elena_ai
TWEET

The ~4MB Rust binary starts instantly inside Termux. I build microservices on the train now.

Kenji Sato
Kenji Sato@kenji_rust
DM
// FOR DEVELOPERS & TEAMS

A plan for every kind of code. Pick yours.

Transparent pricing. No markup. No 5-hour session limits. Cancel anytime. Top up with KBZPay, WaveMoney, or credit cards.

Go Plan

Entry
$1.00USD/ mo
$7.00 in credits included+ $0.35 fee

Micro-entry. Full terminal agent power.

What's included
  • ~20K requests included
  • $7/mo of usage value (7× multiplier)
  • No 5-hour session lockouts
  • 20 req / min rate limit
  • Fusion CLI & Termux mobile support
  • Discord community support
Select Go

Pro Plan

POPULAR
$10.00USD/ mo
$70.00 in credits included+ $0.70 fee

Built for daily builders. Unlocks Cursor & Cline.

What's included
  • ~120K+ requests included
  • $70/mo of usage value (7× multiplier)
  • No 5-hour session lockouts
  • Provider API Unlocked (Cursor, Zed, Cline)
  • 100 req / min priority queue
  • Discord developer support
Select Pro

Max Plan

Team Fleet
$100.00USD/ mo
$300.00 in credits included+ $4.00 fee

High-concurrency pool for software houses & agencies.

What's included
  • ~500K+ requests included
  • $300/mo of usage value (3× multiplier)
  • No 5-hour session lockouts
  • Pooled team bucket (Unlimited seats)
  • 300 req / min enterprise throughput
  • Private 1-on-1 Discord channel
Select Max
// FOR BUILDERS & PREPAID ACCOUNTS

Pay-As-You-Go. The Credit Pack.

No subscription. No commitment. Top up credits that never expire and have zero 5-hour cooldowns.

$1.00USD/ top-up+ $0.35 fee
7× MultiplierNever ExpiresNo Cooldowns

Spend only on the requests you make. Top-up credits roll over into your account and never expire. Works directly with Fusion CLI and the Provider API.

Select Pay-As-You-Go Top-Up
What's included
  • ~20K requests per top-up
  • $7 of usage value per 4.5K MMK top-up (7× multiplier)
  • No Monthly Commitment
  • Credits Never Expire
  • No 5-Hour Session Lockouts
  • Discord community support
Browse Model Rates & Documentation
DeepSeek-V4~150,000+ req / mo
MiniMax-M2.7~45,000+ req / mo
GLM 5.3 Flash (Coming Soon)~180,000+ req / mo
Stripe
Visa
Mastercard
KBZPay
Wave Pay
AYA Pay