Pricing

All features included. Pay only for capacity.

Tendril is BYOK. You pay your AI provider directly at standard rates. Tiers are sized by daily and monthly token usage — pick the headroom you need.

MonthlyAnnualSAVE 26%

Free Tier

$0/month

Free Forever

  • 1M tokens/month, 100k tokens/day- about 15 to 25 pipeline runs per month. Right size for evaluating Tendril or working on one project at a time.
  • Bring Your Own API Key (BYOK)- Anthropic, OpenAI, or Google. You pay your provider directly at standard API rates; Tendril never marks up.
  • Unlimited projects
  • Unlimited parallel agents- capped only by your machine and your provider rate limits.
  • Knowledge Graph + Lessons Learned
  • Live preview + annotation
  • Audit + clickable improvement suggestions
  • GitHub clone + push + open PRs
  • MCP server support
  • Community support
Download Free

Starter Tier

$10/month

Billed annually at $120/year. Save $24.

  • 4M tokens/month, 400k tokens/day- 4× Free headroom. About 60 to 100 pipeline runs per month. Daily project iteration without rationing.
  • Email support
Get Starter - $10/mo

Pro Tier

$20$14/month

Billed annually at $168/year. Save $60.

Most PopularFounding

Highest cap. Same features.

  • 100M tokens/month, 5M tokens/day- 25× Starter monthly headroom. Effectively unlimited for any solo dev workflow.PRO
  • Priority supportPRO
Get Pro - $14/mo
Capacity, not capability

Every tier gets every feature

The difference between Free, Starter, and Pro is how much you can run, not what you can run. Hit the cap, the work waits for the next reset; nothing locks.

Pipeline runs per month

Free: about 15 to 25 runs

Starter: 60 to 100. Pro: 1,500 to 2,500.

Daily token budget

Free: 100k tokens per day

Starter: 400k tokens per day. Pro: 5M tokens per day.

When you hit a cap

Daily: resets at midnight UTC. Monthly: resets on the 1st.

Cap is just usage. Every feature stays available; you can still browse, edit, and review.

Why Tendril beats cold AI

Measured in a 120-trial A/B study

Tendril pipeline orchestration beats raw LLM output at every difficulty level. The advantage grows with task complexity.

+47%Higher quality vs cold OpusMeasured in a 120-run A/B study: Tendril pipeline scored 47% higher on quality overall than a single Claude Opus 4.7 agent working the same tasks (69.2 vs 47.1). Pooled across three task difficulties.
8xGap on hardest taskOn the hardest task in the study (pagination with edge cases plus input validation), Tendril scored 80.5/100 while cold Opus scored 10.2/100. The harder the feature, the bigger the gap.
KGKnowledge Graph compoundingEvery tier primes agents with patterns, memory, and scoped context from past sessions. Planning starts where the last run left off instead of from scratch.
Tendril analytics dashboard showing cost trends and quality scores

Compare Plans

Plans are sized by token usage. Pick the daily and monthly headroom you need.

Feature
Free

$0/mo

Starter

$10/mo

Pro

$14/mo

Tokens per month
1M
4M
100M
Tokens per day
100k
400k
5M
Approx. pipeline runs / month
15-25
60-100
1,500-2,500
Bring Your Own API Key (BYOK)
All pipelines (Simple, Medium, Complex)
Unlimited projects
Unlimited parallel agents
Tendy chat orchestrator
Live preview panel
Knowledge Graph (single-project)
Cross-project Knowledge Graph
Screenshot capture + annotation
Execution levers (patterns, error memory)
Project score + summary on completion
Audit pass with gap-fill retries
Clickable improvement suggestions
GitHub clone + push + PRs
Analytics dashboard
MCP server support
Support
Community
Email
Priority

Frequently Asked Questions

Everything you need to know before getting started.

Install free. Add your own API key and start building.