Identity
- Model
- Ctxdex Frontier 4
- Provider
- Astavra Enterprise AI
- Channel
- Enterprise Managed Ai
- Provider's name for it
- Not recorded
- Offering
- ctxdex:offering:astavra-enterprise-ai:ctxdex-frontier-4
- Currency
- USD
Rates
Charged by Astavra Enterprise AI on this offering. Effective from 15 Sep 2026. Captured as of 14 Sep 2026.
Applies unless another rate set's condition is met.
| Meter | Rate | As recorded |
|---|---|---|
| Audited base input-token rate | $2.10per 1M tokens | No source string recorded |
| Audited base output-token rate | $10.50per 1M tokens | No source string recorded |
| Audited base cached-input-token rate | $0.315per 1M tokens | No source string recorded |
Where this sits in the industry
The external offerings this one is compared against, captured as of 2026-09-14. Each names the surface it is sold on and how well that surface matches this one. No price is shown here: most of these providers do not publish one per unit, and the link goes to whatever they do publish.
Microsoft Foundry — GPT-5.6 Terra
Microsoft Foundry
Sold as Managed Cloud, managed access.
Uses Microsoft Foundry — GPT-5.6 Terra as a verified managed commercial surface for the selected model peer; surface fit to CTXDEX role ENTERPRISE_MANAGED: CLOSE.
This provider prices in its own terms, not captured here. This provider prices in its own terms, which are not captured here.
Provider pricing pageAmazon Bedrock — Claude Sonnet 5
Amazon Bedrock
Sold as Managed Cloud, managed access.
Uses Amazon Bedrock — Claude Sonnet 5 as a verified managed commercial surface for the selected model peer; surface fit to CTXDEX role ENTERPRISE_MANAGED: CLOSE.
This provider prices in its own terms, not captured here. This provider prices in its own terms, which are not captured here.
Provider pricing pageVertex AI — Gemini 3.8 Flash
Google Cloud
Sold as Managed Cloud, public self service.
Uses Vertex AI — Gemini 3.8 Flash as the ranked 3 current commercial/access reference. Exact model availability on a managed cloud surface directly matches the CTXDEX ASTAVRA provider role.
This provider prices in its own terms, not captured here. A cost comparison is possible where the provider's own rate covers the same workload.
Provider pricing pageConditions & rules
A headline rate is not what you pay. Everything below changes it, and none of it is reflected in the figures above.
- Effective from
- 15 Sep 2026
- Effective to
- No end date recorded
- Captured as of
- 14 Sep 2026
- Eligibility
- Published list pricing — available to anyone on this channel.
- Offering status
- Active
- Capacity
- Dedicated Compute, Provisioned Throughput. Capacity beyond shared on-demand is priced on its own basis, shown as its own rate set where one is recorded.
- Commitment
- Enterprise Contract, Term Commitment. The committed amounts are not recorded — a commitment appears here as the discount it carries, not as a price.
- Tier coverage
- 1 tier, each with a recorded cost.
- Separately-priced services
- Map Grounding. These are charged in addition to the rates above.
Pricing rules
A pricing rule is a recorded Commitment — open the definition that changes what a Rate Schedule — open the definition charges — an Allowance — open the definition, a band, a premium or a Spend Commitment — open the definition.
- commitment discount
Eligible charges are 18% less under a recorded commitment.
Pricing versions
Every recorded version of this offering's pricing policy, newest first, with its amounts. A scheduled version is a published future price, not a forecast.
- Astavra Enterprise AIIn forceEffective from 15 Sep 2026 — no end date recordedCaptured as of 14 Sep 2026
- Audited base input-token rate $2.10 per 1M tokens
- Audited base output-token rate $10.50 per 1M tokens
- Audited base cached-input-token rate $0.315 per 1M tokens
Input rate down 13.6% on the previous version.
- Astavra Enterprise AISupersededEffective from 10 Apr 2026 — 15 Sep 2026Captured as of 31 Aug 2026
- Input tokens $2.43 per 1M tokens
- Output tokens $9.72 per 1M tokens
The earliest recorded version, so there is nothing before it to compare with.