The infrastructure behind every call

CallPages runs your voice AI on the best providers in the industry. Bring your own free API keys — when one provider hits a rate limit or quota, your calls automatically fail over to the next one. A call never drops because of a quota.

Create your free call page →I have an account — configure keys

Instant Free Keys (No Credit Card)

Sign up free, grab a key in about 30 seconds, paste it in your dashboard's 🏗️ Infrastructure tab. These providers power voice calls, transcription and AI text — with automatic failover between them.

Google AI Studio (Gemini)🎙️ Recommended for voice

Recommended for voice: Gemini Flash, low latency.

Free: ~15 RPM / 1,500 req/day on Flash (AI Studio free tier; varies by model)
Get a key →

Google AI Studio (Gemini) — vasilyukreino (shared pool)🎙️ Recommended for voice

Shared-pool capacity: full round-robin participant; absorbs voice traffic when the central Gemini key is quota-capped.

Independent AI Studio free tier on the donor account — separate quota from the central key
Get a key →

Google AI Studio (Gemini) — vasilypfknr (shared pool)🎙️ Recommended for voice

Shared-pool capacity: full round-robin participant; absorbs voice traffic when the central Gemini key is quota-capped.

Independent AI Studio free tier on the donor account — separate quota from the central key
Get a key →

Groq🎙️ Recommended for voice

Recommended for voice: Whisper + Llama, low latency.

Free: 30 RPM / 1,000 RPD on Llama 3.3 70B (free tier; see console.groq.com)
Get a key →

OpenRouter

Free: select :free models at ~20 RPM; paid credits otherwise
Get a key →

Cloudflare Workers AI

Free: 10,000 neurons/day (Workers free plan, resets 00:00 UTC)
Get a key →

NVIDIA NIM

Free: limited credits via build.nvidia.com
Get a key →

ElevenLabs (TTS)🎙️ Recommended for voiceIntegration pending

Future voice-synthesis fallback. No server-side TTS call exists in the voice path yet (TTS runs inside the Gemini Live session); key is stored for the future fallback.

Free: 10,000 chars/month TTS (no card; non-commercial)
Get a key →

Cartesia (TTS)🎙️ Recommended for voiceIntegration pending

Future voice-synthesis fallback — sub-90ms streaming TTS. No server-side TTS call exists in the voice path yet (TTS runs inside the Gemini Live session); key is stored for the future fallback.

Free: 20,000 credits/month (~27 min TTS), no card
Get a key →

AssemblyAI (STT)

Free: $50 trial credit at signup
Get a key →

Gladia (STT)

Free: 10 hours/month STT (refreshes monthly)
Get a key →

xAI (Grok)Integration pending

Owner key verified 2026-10-23; no server-side xAI call exists in the voice/text chain yet — config row only.

$25 signup credit reported (unconfirmed in current billing docs — check console); then pay-per-use
Get a key →

Qdrant (vector DB)Integration pending

No retrieval path exists yet — config row only.

Free: 1 GB cluster (Qdrant Cloud free tier)
Get a key →

Pinecone (vector DB)Integration pending

No retrieval path exists yet — config row only.

Free: Starter plan
Get a key →

Fireworks AI (images)

Free: $1 signup credit — FLUX image generation
Get a key →

Hugging Face (images)

Free: Inference API credits on the free tier
Get a key →

Leonardo AI (images)

Free: $5 API credit on signup
Get a key →

DeepAI (images)

Free: limited image generation on the free tier
Get a key →

Zhipu AI (images)

Free: CogView image generation on the free signup package
Get a key →

Trial Credits / Pay-As-You-Go

Generous trial credits or pay-as-you-go — optional extra backups for owners who want even more redundancy. OpenAI and Anthropic have no free API tier; they are supported as paid BYOK options only.

Deepgram (STT)

$200 signup credit (STT)
Get a key →

DeepSeek

Eval grant / low-cost API (see platform.deepseek.com)
Get a key →

AI21 Labs (Jamba)

$10 free credit at signup
Get a key →

Together AI

$25 trial credit at signup
Get a key →

BasetenIntegration pending

Baseten endpoints are per-deployment; add when the owner provides a deployment URL.

$25-30 trial credit at signup
Get a key →

OpenAI (ChatGPT)

Paid only — no free API tier · Paid only — OpenAI offers no free API tier (pay-per-use).
Get a key →

Anthropic (Claude)

Paid only — no free API tier · Paid only — Anthropic offers no free API tier (pay-per-use).
Get a key →

Mistral AI

Free tier on select models; higher tiers need a card
Get a key →

How failover keeps your calls alive

  1. Your key first. Every call tries the provider you configured, with your own API key.
  2. Platform fallback second. If your key hits a limit, CallPages retries with a platform fallback key for the same provider.
  3. Next provider third. Still limited? The call moves to the next provider in your priority order — mid-call, no drop.
  4. Honest when exhausted. If every provider is limited, the call ends with a clear reason — never a silent drop — and your dashboard shows exactly which keys to add.

Put your business on this infrastructure

Get your free CallPages page in 3 minutes: click-to-call AI, your products with descriptions, your own website. Your calls run on the provider matrix above — with automatic failover included.

Create your free call page →