Tandom API
API Pricing
Free up to limits. Pay for what you ship. No surprises.
Free
- Per-endpoint Free monthly limits
- $5/mo AI spend cap per key
- No credit card required
- Community support + docs
Starter
- 10× Free on reference + compute
- 5× Free on AI-bearing endpoints
- $50/mo AI spend cap per key
- Idempotency-Key + webhooks
- Email support
Pro
- 100× Free on reference + compute
- 25× Free on AI-bearing endpoints
- $500/mo AI spend cap per key
- IP allowlisting per key
- Slack support · DPA · 99.9% SLA
Enterprise
- Custom volume + SLA
- DPA + SOC 2 attestation
- Multi-key org accounts
- Bring your own LLM key (optional)
- Dedicated CSM
Per-endpoint quotas
Different endpoints, different cost
HTS lookups are cheap to serve. AI-heavy endpoints (HTS classification, AD/CVD scope analysis, full reports) cost more because they call Claude. Plan multipliers reflect that — reference scales aggressively, AI-heavy scales conservatively.
| Endpoint family | Free | Starter | Pro | Enterprise |
|---|---|---|---|---|
HTS reference search · browse · notes · hierarchy · expand-s301 · expand-ch99-scope | 5,000 calls / mo | 50,000 calls / mo | 500,000 calls / mo | Custom |
Data freshness status | Unlimited | Unlimited | Unlimited | Custom |
Chapter 99 lookup | 2,000 calls / mo | 20,000 calls / mo | 200,000 calls / mo | Custom |
PGA flag check | 2,000 calls / mo | 20,000 calls / mo | 200,000 calls / mo | Custom |
Duty calculation | 1,000 calls / mo | 10,000 calls / mo | 100,000 calls / mo | Custom |
AD/CVD order search | 1,000 calls / mo | 10,000 calls / mo | 100,000 calls / mo | Custom |
AD/CVD shipment check Each lookup fans into 1–10 Haiku scope-analysis calls | 50 lookups / mo | 250 lookups / mo | 1,250 lookups / mo | Custom |
AD/CVD historical rate | 200 lookups / mo | 2,000 lookups / mo | 20,000 lookups / mo | Custom |
AD/CVD full report Scope + lifecycle + rate history per matched case (heaviest fan-out) | 3 reports / mo | 9 reports / mo | 45 reports / mo | Custom |
HTS classification Sonnet-based — closed beta until HTS Classification ships | 25 classifications / mo | 100 classifications / mo | 500 classifications / mo | Custom |
HTS classification report Closed beta | 3 reports / mo | 9 reports / mo | 45 reports / mo | Custom |
MCP tools/list (discovery) | Unlimited | Unlimited | Unlimited | Custom |
MCP tools/call (execution) Each call inherits the underlying tool's limit | Inherits | Inherits | Inherits | Custom |
AI spend cap (per key) Global safety net on top of per-endpoint quotas | $5 / month | $50 / month | $500 / month | Custom |
Rate limit Per-key sliding-window | 60 / minute | 300 / minute | 1,200 / minute | Custom |
Free quotas are locked. Starter / Pro plan prices and per-endpoint multipliers shown here apply when paid tiers go live; existing customers keep their pricing through any future changes.
Plan features
What else you get
| Feature | Free | Starter | Pro | Enterprise |
|---|---|---|---|---|
| Quota reset cadence | Monthly | Monthly | Monthly | Monthly |
| Idempotency-Key + webhooks | — | |||
| Sandbox / test keys | ||||
| IP allowlisting per key | — | — | ||
| API keys per account | 1 | 3 | 10 | Unlimited |
| Support | Community + docs | Priority email + Slack | Dedicated CSM | |
| DPA | — | — | ||
| SLA | — | 99.5% | 99.9% | 99.99% custom |
| SOC 2 attestation | — | — | — | |
| Bring your own LLM key | — | — | — |
How AI cost works
One bill, no surprises
Some Tandom endpoints — HTS classification, AD/CVD scope analysis, classification chat refinement — call Claude under the hood. Tandom pays Anthropic; you pay your plan price.
Each plan absorbs AI cost up to a per-key spend cap (Free: $5 / month · Starter: $50 / month · Pro: $500 / month · Enterprise: custom). When you hit your cap, calls return 429 with Retry-After until reset, OR you raise your cap in your dashboard.
Predictable bills. No per-call AI surcharge. No way for a single bad call — yours or ours — to run up your spend.
Bring your own LLM key is available on Enterprise contracts only — for customers with existing Anthropic agreements + data-residency commitments. Most customers prefer Tandom-managed Claude: faster setup, we handle model selection + quality.
FAQ
Common pricing questions
- Is there really a free tier? No credit card?
- Yes. Sign up with email, verify, get an API key, make calls. No credit card unless you upgrade. The Free tier carries per-endpoint monthly caps + a $5/month Claude-token global safety cap so a single bad call can't run up our Anthropic bill OR yours.
- What happens when I hit my plan limit?
- We return a 429 error with Retry-After. You can wait until reset, raise your spend cap in the dashboard (Starter / Pro / Enterprise), or upgrade your plan. We never let a runaway call surprise-bill you.
- How is AI cost handled?
- Some Tandom endpoints — HTS classification, AD/CVD scope analysis, classification chat refinement — call Claude under the hood. Tandom pays Anthropic; you pay your plan price. Each plan absorbs AI cost up to a per-key spend cap. Predictable bills, no per-call AI surcharge.
- Can I use my own Anthropic key?
- Enterprise tier only. Most customers don't need to — managed Claude is faster to set up and we handle quality + model selection. Some enterprise customers prefer BYOK because they already have an Anthropic enterprise contract with their own DPA + data residency commitments.
- Do you offer SLAs and a DPA?
- DPA + standard SLA on Pro. Custom SLA, SOC 2 audit reports, and contract negotiation on Enterprise.
Start building
Sign up with email. No credit card. ~60 seconds to your first API call.
Free tier is real. No credit card. Browse the API →