Intelliverse Air

Light, fast, impossibly cheap. Ship your first AI feature today.

Quick start

curl https://router-api.intelli-verse-x.ai/v1/chat/completions \
  -H "Authorization: Bearer $INTELLIVERSE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "intelliverse",
    "messages": [{"role": "user", "content": "Hello"}]
  }'

That's the whole integration. No model catalog, no fallback config, no provider juggling — the router classifies each request and targets the best option in your pool.

Context window

32k tokens

Throughput

60 req/min

Included credits

10,000/mo

Media quality

standard

Capabilities unlocked

Fast Chat

High-throughput conversational responses for chat and support workloads.

Code Generation

Code completion, refactoring, and repository-scale reasoning.

Structured Extraction

JSON-mode data extraction, classification, and transformation.

Tool Calling

Reliable function calling for agents and workflows.

Knowledge Base (RAG)

pgvector knowledge base scoped to your App ID, grounded answers with citations.

Monthly credits

10,000 credits granted on every paid invoice — one balance for chat, images, video, voice & 3D. Spend it on any mix below; unused credits roll over and never expire.

10,000

everyday replies

Everyday chat: 1 credits / reply (~1k tokens)

500

standard images

Standard image: 20 credits / image

250

seconds of video

Standard video: 40 credits / second

250

minutes of voiceover

Voiceover: 40 credits / 1k characters (≈1 min narrated)

minutes of music

Music: 720 credits / generated minute

3D assets

3D asset: 250 credits / asset

Each card shows your full monthly grant spent on that single item — in practice you mix freely from one balance (1 credit = $0.001).

Want the full pitch for this plan? See the Air plan page →

Verified pricing

Claude Opus 4 8 (output tokens) on Anthropic, per million tokens: $1.00 (us) vs $25.00 (them)

Firecrawl-checked · we win this row

Read the win fairly: theirs is that one model's per-token rate; ours is the flat Intelliverse Air price for the whole class — the router picks which model serves each request, and it won't always be this one. The row is here for the unit math, not a promise that every request lands on that exact model.

Verified 7/11/2026 · snapshot 4f83a9c253cb… · source

Provider pool: Anthropic, OpenAI, Google, Amazon Bedrock, DeepInfra, Fireworks, Self-hosted vLLM, ElevenLabs, PiAPI, Replicate, Meshy. Which provider serves a given request is decided per-request by the router; model identifiers are masked in response metadata (an opt-in transparency header reports task class, pool, and cost band) — you buy capability, not a model name.