Start Building
Read https://dev-dashboard.floelabs.xyz/agents.md and set up Floe for this project.
Paste into Claude Code, Cursor, or any coding agent.
Works with
Anything that speaks HTTP — every model, every vendor, one key.
The whole bill

One key covers every line on the invoice

Most agent-payment tools stop at LLM tokens. A voice agent's cost isn't tokens — it's transcription, speech, telephony, search, and reasoning across five vendors on a single call. Floe is the one integration in front of all of it, live today.

One voice call · what it actually coststotal $0.0300
TelephonyTwilio
$0.0072
TranscriptionDeepgram
$0.0041
LLM / reasoningOpenAI
$0.0127
SpeechElevenLabs
$0.0038
SearchExa
$0.0022
A token router sees 42%
Just the LLM slice — $0.0127 of the call.
Floe sees 100%
Every vendor, every line — $0.0300.
LayerVendorsWhat's live
LLM / reasoningOpenAI, Anthropic, Venice AIPay per call, any model, via one Floe key
Speech (STT / TTS)Deepgram, ElevenLabsSame key, same ledger, sub-cent per call
TelephonyFloe Phone via TwilioOn the same ledger + spend caps**
Memory / data / searchHydraDB, Exa, TavilyMetered per call, allowlisted per agent

** Floe Phone spend is unified onto the same ledger with the same spend caps.

Token routers like OpenRouter meter LLM spend. Floe meters the entire agent bill — speech, telephony, search, compute — so your budgets, caps, and cost analytics cover 100% of what a call costs, not the fraction that's tokens.

Integrate in minutes

Already have a pipeline? Keep it.

Start Building
from pipecat_floe import FloeSTTService, FloeLLMService, FloeTTSService

stt = FloeSTTService()                            # streaming STT
llm = FloeLLMService(model="openai/gpt-4o-mini")  # OpenAI-compatible
tts = FloeTTSService(model="openai/tts-1", voice="alloy")
# drop into your Pipeline as usual — one cap bounds the whole run
pip install pipecat-floe — all three legs, one key, one budget. Verified against pipecat-ai 1.7.0.
Integrate in minutes

Already have a pipeline? Keep it.

Start Building
from pipecat_floe import FloeSTTService, FloeLLMService, FloeTTSService

stt = FloeSTTService()                            # streaming STT
llm = FloeLLMService(model="openai/gpt-4o-mini")  # OpenAI-compatible
tts = FloeTTSService(model="openai/tts-1", voice="alloy")
# drop into your Pipeline as usual — one cap bounds the whole run
pip install pipecat-floe — all three legs, one key, one budget. Verified against pipecat-ai 1.7.0.
For production teams

Spending $1k+/mo across your voice stack?

15 minutes with the founders. Bring your vendor list — we'll map your whole bill onto one ledger on the call.

Alex Christian — Co-founder, Floe
Book a demo
Keyless LLM · one key

Every model your agent needs — no per-provider accounts

Point your OpenAI-compatible SDK at Floe, pass your Floe key, and call models pay-as-you-go. No separate account and key for every provider — Floe fronts the vendor and meters it on your one ledger. Fund with a card; your agent never touches a wallet.

floe_•••key
Open-weight · live & keyless
LlamaQwenDeepSeekMistralGemmaCommand-R

Served keyless via the Floe gateway, pay-as-you-go — self-hosted and serverless, so no per-provider key to manage.

BYOK optional. Bring your own provider key and pay at cost, or go keyless and let Floe front it.
Open source

Start with the guardrail. No account.

floe-guard hard-stops a runaway agent before it burns your bill. Local, framework-agnostic, MIT.

$pip install floe-guard
Need caps enforced server-side, across vendors, on one ledger?Start Building
How it works · The full financial loop
Pay any vendor per call. Authorize every dollar. Bill it from one place.
Connect a card and your agent can pay thousands of vendors via pay as you go APIs with hard spend authorization enforced before the money moves — and one ledger across every vendor.
Step 1 · Fund
Connect a card — your agent is funded
Top up from a card or bank in 100+ countries. Visa, Mastercard, Apple Pay, Google Pay.
Visa · Mastercard · Apple/Google Pay · 100+ countries
// USDC liquidity in. Global access out.

  IN  Agent Wallet    OUT
  ────             ────
  Cards           Virtual cards
  Bank transfer    Bank accounts
  Apple Pay       Local payouts
  Google Pay      On-chain
Step 2 · Pay
Pay as you go any vendor API, per call — Floe settles the vendor
Send any URL through the Floe proxy. Floe pays the vendor, settles instantly, and debits your agent’s ledger — down to sub-cent, with no 30¢ card floor. The agent just gets its response back; it never touches the payment layer.
2,000+ payable APIs* · ~50ms signing
// Agent sends the target URL - that's it
await fetch("https://credit-api.floelabs.xyz/v1/proxy/fetch", {
  method: "POST",
  headers: { "Authorization": `Bearer ${apiKey}` },
  body: JSON.stringify({
    url: "https://api.anthropic.com/v1/messages",
    method: "POST",
    body: payload
  })
});

// → Floe settles the vendor, debits your ledger,
//   returns the response. No wallet, no SDK.
// Response header: X-Floe-Payment-Amount
Step 3 · Authorize
Set spend authorization it can’t blow past
Hard limits per agent, per vendor, per task — optionally time-bound — enforced before the call, not on next month’s invoice. Stop a runaway agent at $1 instead of discovering a $400 loop after the fact.
Pre-call enforcement · per-agent · per-vendor · time-bound
// Authorize spend at the agent + vendor level
await fetch("https://credit-api.floelabs.xyz/v1/spend-controls", {
  method: "POST",
  body: JSON.stringify({
    agentId: "agt_research_01",
    dailyCapUsd: 50,
    perCallCapUsd: 2,
    vendors: ["anthropic", "pinecone"], // allowlist
    expiresAt: "2026-06-30T00:00:00Z"
  })
});

// Any call over the cap or off the allowlist is
// rejected BEFORE money moves - a 402, not a bill.
Step 4 · Bill
One ledger across every vendor
Every payment — across all your APIs and all your agents — lands on one itemized ledger. Per-call cost, per-vendor totals, per-agent and per-customer attribution. The spend visibility card-based and per-provider billing can’t give you.
Per-call · per-vendor · per-agent · per-customer
// Unified spend ledger, filterable by agent
const feed = await fetch(
  "https://credit-api.floelabs.xyz/v1/developer/activity?agentId=agt_research_01"
)

// Returns an itemized, unified stream:
// [
//   { vendor: "anthropic", call: "messages",
//     costUsd: 0.012, ts, agentId, customerId },
//   { vendor: "pinecone",  call: "query",
//     costUsd: 0.003, ts, agentId, customerId },
//   ...
// ]

// Roll up by vendor, agent, or customer to bill.

"Your voice agent pays only for the tools and models it needs. Per call. In real time. That is a 10x reduction in API procurement overhead for high-volume teams. And a story that sells itself to enterprise security buyers."

Muhammad Bilal

SOC Voice Developer

1/3

"Floe's voice AI infrastructure solves one of the missing pieces that voice AI builders need to operate autonomously in the real world: multi vendor billing and secure spending governance."

Ratikesh

Voice AI Developer at Pharma Company

2/3

"Floe is solving the exact friction point I’ve hit while building production voice agent architectures: one key for any model and tool."

Ali

AI Voice Agent Developer

3/3

"Your voice agent pays only for the tools and models it needs. Per call. In real time. That is a 10x reduction in API procurement overhead for high-volume teams. And a story that sells itself to enterprise security buyers."

Muhammad Bilal

SOC Voice Developer

1/3

"Floe's voice AI infrastructure solves one of the missing pieces that voice AI builders need to operate autonomously in the real world: multi vendor billing and secure spending governance."

Ratikesh

Voice AI Developer at Pharma Company

2/3

"Floe is solving the exact friction point I’ve hit while building production voice agent architectures: one key for any model and tool."

Ali

AI Voice Agent Developer

3/3

Pricing

Free to integrate and try. 5% per spend through Floe.

$0to integrate
No subscription
No minimums
No seat fees
Get Started
Meet the team

Who's behind Floe

Floe Labs founders
Alex Christian
Co-founded DataMynt. Former Airwallex, Western Union, eBay.
Shivam Chaturvedi
Co-founded DataMynt. Previously CTO Kado.money (Acquired), Founding engineer at Transak.

Book a demo with the founders
15 minutes. We'll map your vendor bill on the call.
FAQ
Your questions, answered

Your agent calls APIs, and many of them charge per call — OpenAI, Claude, Kimi, Deepgram, ElevenLabs, Twilio, search, memory. Floe pays each vendor per call through one key, on one ledger. You authorize every dollar before it moves.

One integration replaces the billing stack — funding, per-call payment across every vendor, spend controls, a unified ledger, and routing to tools and models you need for voice AI.

One Floe key covers your whole stack, live today: LLMs (OpenAI, Anthropic, Venice AI), speech (Deepgram, ElevenLabs), telephony (Twilio), and memory, data & search (HydraDB, Exa, Tavily) — part of a marketplace of 36+ services across 8 categories.

Open-weight models — Llama, Qwen, DeepSeek, Mistral, Gemma, Command-R — run keyless via the Floe gateway, pay-as-you-go, with no per-provider account to manage. Point your OpenAI-compatible SDK at Floe, pass your Floe key, and call. Browse the marketplace →

Token routers meter your LLM spend. For a voice agent, tokens are only a fraction of the bill — the rest is transcription, speech, telephony, and search, spread across several vendors on a single call.

Floe meters the entire agent bill, not just tokens. That’s why your budgets, caps, and cost analytics cover 100% of what a call costs — and why Floe can actually bound what an agent spends, rather than watching one slice of it. Floe sits as the governance layer above your whole stack, token routers included.

One install, one proxy endpoint. Drop-in SDKs for AgentKit, LangChain, CrewAI, Vercel AI SDK, OpenAI Agents, and Claude (MCP). Or hit the REST API directly from Python, Rust, Go — anything that speaks HTTP. Funds settle in stablecoins under the hood, but you never touch a wallet or a seed phrase.

Full quickstart: Agent Quickstart →

Funding via Visa, Mastercard, ACH, Apple Pay, Google Pay, or local payouts in 100+ countries. Money settles into your agent’s wallet.

Pricing: Free to integrate — no subscription, no seat fees, no minimums. Floe earns 5% fee on the volume that flows through it; the vendors you call charge their own rates.

Spend caps are enforced before money moves, not after. You set per-call and daily limits, allowed vendors, rate ceilings, and time-bound permissions in code. Any call over a cap or off the allowlist is rejected before payment — a 402, not a surprise on next month’s invoice. Plus a self-serve kill-switch: pause any agent instantly, or let a policy auto-trip it. Stop a runaway loop at $1 instead of finding a $400 one after the fact.

Floe is built by Alex Christian (DataMynt co-founder; Airwallex, Western Union, eBay) and Shivam Chaturvedi (DataMynt co-founder; CTO at Kaido Money, acquired; founding engineer at Transak). Both shipped payments, compliance and billing infrastructure at scale before Floe.

See example agents on GitHub →

Or book a demo with the founders directly →

FAQ
Your questions, answered

Your agent calls APIs, and many of them charge per call — OpenAI, Claude, Kimi, Deepgram, ElevenLabs, Twilio, search, memory. Floe pays each vendor per call through one key, on one ledger. You authorize every dollar before it moves.

One integration replaces the billing stack — funding, per-call payment across every vendor, spend controls, a unified ledger, and routing to tools and models you need for voice AI.

One Floe key covers your whole stack, live today: LLMs (OpenAI, Anthropic, Venice AI), speech (Deepgram, ElevenLabs), telephony (Twilio), and memory, data & search (HydraDB, Exa, Tavily) — part of a marketplace of 36+ services across 8 categories.

Open-weight models — Llama, Qwen, DeepSeek, Mistral, Gemma, Command-R — run keyless via the Floe gateway, pay-as-you-go, with no per-provider account to manage. Point your OpenAI-compatible SDK at Floe, pass your Floe key, and call. Browse the marketplace →

Token routers meter your LLM spend. For a voice agent, tokens are only a fraction of the bill — the rest is transcription, speech, telephony, and search, spread across several vendors on a single call.

Floe meters the entire agent bill, not just tokens. That’s why your budgets, caps, and cost analytics cover 100% of what a call costs — and why Floe can actually bound what an agent spends, rather than watching one slice of it. Floe sits as the governance layer above your whole stack, token routers included.

One install, one proxy endpoint. Drop-in SDKs for AgentKit, LangChain, CrewAI, Vercel AI SDK, OpenAI Agents, and Claude (MCP). Or hit the REST API directly from Python, Rust, Go — anything that speaks HTTP. Funds settle in stablecoins under the hood, but you never touch a wallet or a seed phrase.

Full quickstart: Agent Quickstart →

Funding via Visa, Mastercard, ACH, Apple Pay, Google Pay, or local payouts in 100+ countries. Money settles into your agent’s wallet.

Pricing: Free to integrate — no subscription, no seat fees, no minimums. Floe earns 5% fee on the volume that flows through it; the vendors you call charge their own rates.

Spend caps are enforced before money moves, not after. You set per-call and daily limits, allowed vendors, rate ceilings, and time-bound permissions in code. Any call over a cap or off the allowlist is rejected before payment — a 402, not a surprise on next month’s invoice. Plus a self-serve kill-switch: pause any agent instantly, or let a policy auto-trip it. Stop a runaway loop at $1 instead of finding a $400 one after the fact.

Floe is built by Alex Christian (DataMynt co-founder; Airwallex, Western Union, eBay) and Shivam Chaturvedi (DataMynt co-founder; CTO at Kaido Money, acquired; founding engineer at Transak). Both shipped payments, compliance and billing infrastructure at scale before Floe.

See example agents on GitHub →

Or book a demo with the founders directly →