Capabilities your agents can call
ASR, document parsing, image generation, TTS, real-time translation, crawling — every capability backed by multiple providers, gated per agent, billed in one place.
Everyday capabilities, already wired up
Voice, documents, images — the things agents need at work, integrated for you. Each capability speaks to multiple providers, so you can switch on quality, price or compliance without writing adapters.
Speech recognition
Transcribe audio with multi-language, long-form and speaker diarization support.
ElevenLabsText to speech
Read agent output aloud — multi-language voices for podcasts, videos, narration.
ElevenLabs
Google GeminiDocument parsing
PDF, Word, Excel, PPT, images — extract text, tables and layout while preserving structure.
Reducto
UnstructuredImage generation
Describe what you need in chat — agents produce illustrations, covers, banners, vectors.
OpenAIGoogle GeminiWeb search · Deep research
Search the web, pull verified sources, run multi-step research — the way agents do research, comparisons, and reports faster.
Perplexity
ExaGoogle GeminiWeb crawling
Bulk crawling with dedup, cleanup, table conversion — the workhorse for data research.
CloudflareToggle per agent, decide who can use what
You decide which capabilities each agent can reach — Researcher doesn't need image generation, Writer does. This same toggle drives quota, usage stats, and risk isolation.
RResearcher
Speech recognition
Text to speech
Document parsing
Image generation
WWriter
Speech recognition
Text to speech
Document parsing
Image generation
One invoice, every provider rolled up
All capability calls aggregate into a single monthly bill — no need to manage each vendor's billing. Usage breakdowns by agent and capability, with auto-alerts when a budget is hit.
Usage · April 2026$135.65
Speech recognition8,420 calls$48.20
Text to speech5,103 calls$32.10
Document parsing1,872 calls$24.90
Image generation4,931 calls$18.40
Web search · Deep research2,118 calls$12.05
Turn on capabilities for your agents
Sign in and grant the capabilities you need — minutes to integrate, no contracts to sign, no APIs to wire up.