Requesty vs. AssemblyAI

Learn why developers building Voice AI choose AssemblyAI’s LLM Gateway over Requesty:

0% markup—no 5% fee added to every call
Native speech-to-text in the same pipeline—Requesty routes text only
Automatic fallbacks, EU data residency by default, and opt-in zero data retention

Get your API key See the comparison

At a glance: Requesty vs. AssemblyAI

Model

AssemblyAI LLM Gateway

Requesty

Pricing

0% markup

5% markup on every call

Models

25+ curated production models

500+ models, 30 providers

Speech-to-text

First-party models—no extra hop

—

Automatic fallbacks

Configurable per-model

Fallback chains

EU data residency

Default-on for EU traffic

Frankfurt region

Zero data retention

Opt-in, per request

Default

SOC 2 Type 2

In progress

BAA available

—

OpenAI SDK compatible

Everything you get with the AssemblyAI LLM Gateway

0% markup

Pay provider token rates with no gateway markup, no per-call surcharge, and no minimum commitment—not a 5% tax that grows with every request.

Automatic fallbacks

Configure a primary model and any number of backups per request. If a provider errors or rate-limits, the Gateway retries the next model—no code changes.

Native speech-to-text

Transcribe with AssemblyAI’s own speech models and apply any LLM to the result in one pipeline—no second vendor and no extra network hop.

25+ frontier models

GPT, Claude, Gemini, Qwen, Mistral, and more behind one OpenAI-compatible API. New models are added the day they launch.

OpenAI-compatible

Drop into any OpenAI SDK or HTTP client. Change a base URL and a model string, and the rest of your code keeps working.

EU data residency

Route requests through EU-resident infrastructure, on by default for traffic originating in the EU, for GDPR-sensitive workloads.

Zero data retention

Opt into zero data retention per request or project-wide, so prompts and responses are never stored.

BAA available

Sign a Business Associate Addendum for applications that process PHI, backed by SOC 2 Type 2, ISO 27001, PCI DSS, and GDPR.

Start building

Get your free API key and make your first Gateway call in minutes—no commitments or minimums.

Live demo

Pick a model. Take action on your audio.

The same transcript, four different jobs, four different models — all routed through one endpoint.

Prompt

What is runner's knee?

claude-opus-4-7 412 ms · 47 tokens

Based on the transcript, runner's knee is a condition characterized by pain behind or around the kneecap. It is caused by overuse, muscle imbalance and inadequate stretching. Symptoms include pain under or around the kneecap and pain when walking.

Ready to drop the 5% markup?

Move to one OpenAI-compatible API with 0% markup, automatic fallbacks, and native speech-to-text—no per-call fee.

Get your API key

Playground

We’re not playing around—but you can

Put our Voice AI models and the LLM Gateway to the test in our no-code playground.

Explore Playground

Frequently asked questions

: Both are hosted, OpenAI-compatible gateways with per-model fallbacks, but AssemblyAI charges 0% markup and adds native speech-to-text in the same pipeline, while Requesty adds a 5% markup on every call and routes text only. Requesty’s strength is cost-routing across 500+ models; AssemblyAI’s is no-markup pricing plus first-party audio. Choose AssemblyAI when you’re building on voice and want to avoid a per-call fee.
: Yes, on gateway fees. AssemblyAI passes through provider token rates with 0% markup, while Requesty adds 5% on top of every model’s base cost—a fee that scales linearly with usage. Requesty can offset some of that with smart routing to cheaper equivalent models, but the 5% applies to every call.
: AssemblyAI, because it runs its own speech-to-text models and applies any LLM to the transcript in one pipeline—no second vendor and no extra network hop. Requesty has no native speech capability; any audio would be a passthrough to a third-party model.
: Yes. Both support per-request fallback chains, and both offer EU data residency (AssemblyAI default-on for EU traffic; Requesty via its Frankfurt region). Requesty applies zero data retention by default, while AssemblyAI makes ZDR an explicit per-request or project-wide opt-in.
: With AssemblyAI, yes—it is SOC 2 Type 2 certified and can sign a Business Associate Addendum (BAA) for workloads that process PHI. Requesty’s SOC 2 Type 2 is still in progress and it does not document a BAA, which is a blocker for regulated healthcare use.
: Requesty is a strong fit if your priority is maximum model breadth (500+ models across 30 providers) with automatic cost-routing and semantic caching to minimize spend. AssemblyAI is the better fit when you want no per-call markup, native speech-to-text, and compliance you can sign today.