Skip to main content

Talonic Pricing

Ingest documents at 0.05 per page. Registry queries are free forever. Four tiers: Free (0/month, 5,000 credits), Build (49/month, 65,000 credits, 25% off), Scale (499/month, 750,000 credits, 33% off), and Enterprise (custom volume pricing with SLA). All tiers include the full extraction pipeline, Field Registry, schema validation, OCR, provenance, and webhook delivery. Credits are the universal unit: 1,000 credits = 1. Page ingestion costs 50 credits per page. Structuring cells cost 20 credits each (registry-resolved cells are free). Intelligence operations cost 100 credits each. Batch mode runs at 0.5x cost with 48-hour delivery.

PRICING

Ingest once. Query forever.

0.05 per page. Registry queries free. Pay only for what your AI agents actually consume.

Free

€0/ month

5,000 credits / month

5 usage value

Best for: Prototyping, agent eval

  • Full 4-phase pipeline
  • Field Registry with savings
  • All 25+ file formats
  • Hard cap — no overage
  • Community support
Sign up free →
RECOMMENDED

Build

€49/ month

65,000 credits / month

65 usage value

25% off à-la-carte

Best for: Solo devs, single-agent workflows

  • Everything in Free
  • Auto top-up on overage
  • Top-up packs unlock
  • Email support
  • 99.5% uptime
Talk to sales →

Scale

€499/ month

750,000 credits / month

750 usage value

33% off à-la-carte

Best for: Teams, multi-tenant agents

  • Everything in Build
  • Multi-user organizations
  • Role-based access
  • Priority support
  • 99.9% uptime
Talk to sales →

Enterprise

Custom

Custom commit

Volume + SLA pricing

Best for: SSO, ISO 27001, BAA, SLA, on-prem

  • Everything in Scale
  • SSO / SAML
  • ISO 27001 / HIPAA / BAA
  • On-prem deployment
  • Dedicated CSM
  • Custom DPA
Talk to sales →

Estimate your monthly cost

Move the sliders. Watch the Registry savings grow as you ingest more.

135.00

135,000 credits

Registry savings

28.50 saved by your graph this month

Recommended tier

Scale

Includes 750,000 credits/mo

Start with Scale →

Compare against doing it yourself.

Move the sliders. See how the registry flattens extraction cost as systems multiply.

Without Talonic

€4,000 / month

20,000 pages × 5 systems × $0.0043 raw API × 10× production overhead

Includes engineering, OCR for scans, validation, retries, infrastructure.

With Talonic

€1,000 / month

1,000,000 credits · all systems included

Registry, schema validation, OCR, provenance, delivery — included.

Monthly savings with Talonic

€3,000 / month

€36,020 / year

Book a Demo

Raw API cost based on Claude Sonnet 4.6 public pricing: $3.00/M input tokens, $15.00/M output tokens (verified May 2026). Average page modeled at 700 input tokens + 150 output tokens.

The 10× production overhead multiplier reflects realistic DIY costs beyond raw API tokens: engineering time (1-2 FTE for production extraction pipeline), OCR for scanned and handwritten content, schema validation, retry and error handling, observability, and infrastructure. Actual multipliers vary by volume — typically higher (15-50×) at moderate volumes, lower (3-10×) at very high volumes.

Per-operation pricing

Every operation in Talonic, billed in credits and EUR.

OperationDescriptionEURCredits
Page ingest (OCR + Registry contribution)per page€0.05050
Structuring cell (AI-resolved)per cell€0.02020
Structuring cell (Registry-resolved)per cellFree0
Intelligence operation (Matching, Cases)per op€0.100100
Markdown-only ingest (no Registry contribution)per page€0.01515
Batch mode (applied to Ingest + Structuring)multiplier0.5×0.5×

1,000 credits = 1.00. Every API response includes credit-cost and balance headers.

Top-up packs

Starter

10,000

credits

10

Builder

55,000

credits

50

+10%

Pro

287,500

credits

250

+15%

Scale

1,250,000

credits

1,000

+25%

Top-up credits stack on tier allowances and roll forward 12 months. Tier credits reset monthly.

The Registry advantage

Why Talonic gets cheaper the more you use it.

Most extraction APIs charge you full price for every document, every time. Talonic doesnt. Every page you ingest contributes fields to your Field Registry a knowledge graph of canonical fields specific to your data.

When you run a future job, ~30% of cells resolve directly from the graph with zero AI calls. Those cells cost you nothing. As your Registry grows, that ratio climbs.

Pay once to ingest. Query forever, asymptotically free.

30%+ cells resolved free, on average

For AI agents

MAKE A REQUEST

curl -X POST https://api.talonic.com/v1/extract \
  -H "Authorization: Bearer $TALONIC_API_KEY" \
  -F "file=@invoice.pdf" \
  -F 'schema={"vendor_name":"string","total_amount":"number","due_date":"date"}'

READ THE COST

HTTP/1.1 200 OK
X-Talonic-Cost-Credits: 70
X-Talonic-Cost-EUR: 0.07
X-Talonic-Balance-Credits: 64930
X-Talonic-Cells-Resolved-Registry: 0
X-Talonic-Cells-Resolved-AI: 1

CHECK BALANCE

GET https://api.talonic.com/v1/credits/balance

{
  "balance_credits": 64930,
  "balance_eur": 64.93,
  "burn_rate_30d_credits": 12400,
  "projected_runway_days": 157,
  "tier": "pro",
  "tier_resets_at": "2026-05-01T00:00:00Z"
}

Auto top-up is human-gated: a human enables it via PATCH /v1/billing/settings, then agents can call POST /v1/billing/topup. The agent cannot enable top-up itself (returns 403 if not enabled).

Frequently asked questions

What is a credit?
One credit equals €0.001. Every operation is metered in credits — but every price is also published in EUR so you always know what you’re paying.
What does “Registry queries are free” actually mean?
Every page you ingest contributes fields to your Field Registry — a knowledge graph of canonical fields specific to your data. Future jobs resolve ~30% of cells (and growing) directly from the graph with zero AI calls. Those cells cost nothing.
What happens when I exceed my tier allowance?
Build and Scale tiers auto top-up at the same per-operation rate (Builder pack increments — €50 for 55,000 credits +10%). Free tier is a hard cap; the API returns 429 until the next month or you upgrade.
Do unused credits roll over?
Tier credits reset monthly. Top-up credits roll forward for 12 months from purchase.
How does batch mode work?
Set processing_mode=batch on upload. Stage 1 (OCR + classification) runs immediately. Stage 2 (extraction) runs against the provider’s batch API with a 48-hour delivery window. Bills at 0.5× on Ingest and Structuring.
Can my AI agent buy credits autonomously?
Top-up is human-gated. A human enables auto top-up via PATCH /v1/billing/settings, then the agent can call POST /v1/billing/topup programmatically. The agent cannot enable top-up itself (403 if not enabled). Every API response carries credit-cost and balance headers.

Ingest once. Query forever.

Start free