Managed · Cloud
Pay per model, in credits.
Load a credit balance, then call any model with one token. Each request draws down at that model's Yantra rate — 50–90% below the commercial equivalent. No per-model subscriptions, no idle seat fees, no surprise invoices.
How it works
Credits in, usage out.
Top up a shared org balance from $250 up — bigger packs carry bonus credits. Credits never expire while your account is active.
One token reaches chat, image, video, speech, avatar, and vision. Set per-token rate limits so teams can't overspend.
Every request is metered at its exact unit rate and logged. Auto top-up keeps the balance above your threshold.
Credit packs
The more you load, the more you get.
1 credit = $1 of metered usage at the rates below. Volume bonus is added on top.
Rate card
What each model costs — vs commercial.
Live rates from the platform's pricing engine. The commercial column is the verified rate for the nearest equivalent from OpenAI, Anthropic, DALL·E, HeyGen, Runway, or AWS.
| Model | Unit | Yantra | Commercial | You save |
|---|---|---|---|---|
| Chat & code · per 1M tokens (in / out) | ||||
| gemma3:1b | tokens | $0.02 / $0.06 | $0.05 / $0.20 | 70% |
| gemma4:8b | tokens | $0.05 / $0.20 | $0.10 / $0.40 | 50% |
| gpt-oss:20b | tokens | $0.10 / $0.30 | $0.15 / $0.60 | 50% |
| qwen3.6:36b | tokens | $0.30 / $1.00 | $0.80 / $4.00 | 73% |
| qwen3-coder:480b :cloud | tokens | $1.20 / $4.00 | $3.00 / $15.00 | 72% |
| Image · per image (1024²) | ||||
| flux-schnell | image | $0.001 | $0.003 | 67% |
| sdxl | image | $0.003 | $0.040 | 92% |
| flux-dev | image | $0.005 | $0.055 | 91% |
| Speech · TTS per char · STT per clip | ||||
| kokoro (TTS) | character | $0.0000075 | $0.000015 | 50% |
| whisper (STT) | per clip | ~$0.003 | ~$0.006 | 50% |
| Video · per clip | ||||
| ltx | per clip | $0.05 | $0.10 | 50% |
| wan2.2 | per clip | $0.15 | $0.30 | 50% |
| Avatar · talking-head · per clip | ||||
| wav2lip | per clip | $0.25 | $0.50 | 50% |
| sonic | per clip | $0.30 | $0.60 | 50% |
| Vision · per image | ||||
| yolo11-seg | image | $0.0003 | $0.003 | 90% |
| sam2.1 | image | $0.0005 | $0.005 | 90% |
Local models are billed on compute only; :cloud models pass through frontier providers and are the one case where data leaves the network.
Worked example
A month of real usage.
Mid-size product team
Layer organization controls on top of credits.
- Enforced SSO & SCIM provisioning
- Priority routing & higher rate caps
- Usage analytics & per-team budgets
- Monthly invoicing & PO billing
Start small, scale later
Load a pack and build today.
Pilot with a starter balance; we'll help you right-size once usage settles.