Get 2,500 events tracked for freeSign up now

Your Agentic AI budget should be a control, not a report.

Give every agent an identity, a budget, and a hard limit enforced before the provider call. One line of code.

Try for free

2,500 events per month, free

Live spendRealtime
summarizer_agent
$431
reasoning_agent
$187
execution_agent
$24
Budget used18%

Works with the providers your agents already call

  • OpenAI
  • Anthropic
  • Gemini
  • DeepSeek
  • Kimi

To your provider, your whole system is one API key and one monthly total.

Account-level caps fail in multi-agent systems because several agents share one credential, so hitting the cap stops all of them rather than the one misbehaving. Provider limits are commonly notification-only, and by the time an alert is read the spend already exists.

Capsera sees each agent and each run separately, and refuses the call before the money leaves.

Anthropic

Invoice

Invoice numberANT-2026-07-4471
Billing period1–31 July 2026
Accountacct_9f2c41
DESCRIPTIONAMOUNT
API usage$642.00

No further breakdown available.

Total due$642.00

Capsera

Invoice · same period

11 agents
Invoice numberCAP-2026-07-0118
Billing period1–31 July 2026
Accountacct_9f2c41
AGENTBUDGETSPEND
Anthropic$642.00
  • reasoning_agent$460$301.40
  • data_extractor$500$164.20
  • support_bot$95$88.15
  • summarizer_agent$150$31.60
  • email_drafter$140$18.90
  • doc_indexer$100$12.75
  • execution_agent$80$9.40
  • intent_router$60$6.30
  • lead_scorer$50$4.55
  • qa_checker$40$3.20
  • report_builder$25$1.55
Total due$642.00

Your AI bill should never surprise you

Same month. Same agents. One has a ceiling.

Without Capsera

Budget

$22,635

The invoice arrives after the damage is done.

With Capsera

Budget

$1,000

The call is blocked before you are billed.

How it works

From one anonymous API key to a governed fleet.

  1. 01

    Install the SDK

    One line at the top of your app. No proxy, no gateway, and no changes to your existing call sites.

    pip install capsera
  2. 02

    Give every agent an identity

    Sub-agents inherit their parent's identity from an in-process context stack, so a graph composed at runtime is attributable without threading a tag through every request.

    capsera.init(agent="data_extractor")
  3. 03

    Set the envelope

    A budget per agent and per run, with a block action rather than an alert. The limit is checked before the provider call, not reconciled after the invoice.

    capsera.budget(monthly=500, on_exceed='block')
  4. 04

    Watch it, and route around it

    Spend attributed per agent, per run and per customer, with cost-aware routing when a cheaper model would do.

    capsera.route(prefer='cost')
  5. 05

    Get told before it mattersComing

    A threshold on the envelope, not just the ceiling. The warning arrives while there is still something to decide, rather than as a line on next month's invoice.

    9:41
    Capsera Alertsnow

    reasoning_agent is at 66% of its monthly budget

    $301.40 of $460. At 100% the next call is refused rather than billed.

  6. 06

    Ask from where you already workComing

    An MCP server, so the assistant in your editor can answer what an agent cost, what is near its budget and what to route differently, without leaving the thing you were doing.

    ClaudeCursorVS Code, and others
    mcp.jsonIllustrative
    {
      "mcpServers": {
        "capsera": {
          "command": "capsera-mcp"
        }
      }
    }

    Shape only. The published command and its options will land with the server.

Overview

Org-wide spend, agent activity, and governance status at a glance.

⛉ GovernanceBudgets2 on trackAlerts0 open

Total spend

$1.69

Last 7 days

Total calls

200

API requests

Active agents

4

Unique agents

Tokens used

514.8K

344.8K in · 170.0K out

Daily spend

8 days · last 7 days

$0.600$0.450$0.300$0.150$0.000
Jul 22Jul 23Jul 24Jul 25Jul 26Jul 27Jul 28Jul 29

Top agents

View all ›

Ranked by spend this period

  • 1reasoning_agent$1.2743 calls
  • 2data_extractor$0.21460 calls
  • 3it_support_agent$0.10444 calls
  • 4summarizer_agent$0.08050 calls

Model breakdown

Analyze ›

Spend by model

  • claude-sonnet-4-6$1.55
    60 calls · anthropic92% of spend
  • claude-haiku-4-5-20251001$0.089
    80 calls · anthropic5% of spend
  • gpt-4o-mini$0.042
    60 calls · openai2% of spend

Live events

Live

Waiting for events. Traffic appears here in real time.

Works with

Thirteen providers and the frameworks you already build on.

The SDK patches the provider clients in place at init, sync and async, so nothing about how you make calls has to change.

Model APIs

Called directly from your own code.

  • AnthropicClaude
  • OpenAIGPT
  • GoogleGemini
  • MistralMistral, Codestral
  • CohereCommand, Embed
  • DeepSeekDeepSeek-V3, R1
  • KimiMoonshot

Inference platforms

Open models, served fast.

  • GroqLPU inference
  • TogetherOpen model hosting
  • FireworksOpen model hosting
  • CerebrasWafer-scale inference

Cloud model services

Behind your existing cloud account.

  • Vertex AIGoogle Cloud
  • BedrockAWS

Agent frameworks

Attribution follows the graph, not the call site.

  • LangGraphStateful agent graphs
  • CrewAIMulti-agent crews
  • LangChainChains and tools
  • LlamaIndexRetrieval pipelines
  • AutoGenConversational agents

Against a provider dashboard

Six things a provider dashboard cannot tell you.

  • Spend per agent or feature
  • Spend per end customer
  • Gross margin per customer
  • Per-agent budget envelopes
  • Hard limits enforced before the call
  • Cost-aware model routing

Every row above is something Capsera does and a provider dashboard does not. Setup is one line of code either way.

LLM providers
Usage
Billing period1–31 July 2026

Total usage

$642.00

1 Jul16 Jul31 Jul
API usage$642.00

No further breakdown available. Usage is reported per API key.

Capsera
Capsera
1–31 July 2026

Total spend

$642.00

Eleven agents, each with its own line.

reasoning_agentof $460$301.40
data_extractorof $500$164.20
support_botof $95$88.15
summarizer_agentof $150$31.60
email_drafterof $140$18.90
doc_indexerof $100$12.75

A provider usage console. Illustration, not a screenshot. Generic on purpose, because the limitation is in the data model rather than any one vendor’s design.

The same month in Capsera. Same total, four ways of asking what it was spent on.

Primary research · 2026-07-31

We did not assume the problem. We measured it.

Cost defect

Clean

39 of the 51 repositories making direct LLM calls carried at least one cost defect. Not bugs: the code works. Patterns that quietly multiply cost, and none of them show up on a provider dashboard.

10,350
files scanned
across 133 repositories
10%
run frontier models in tests
10% (5 of 51), 28 instances
293
agents found, median 2 per repo
293 total, median 2 per repo
62%
hide calls behind a framework
62% (82 of 133)
Read the method and limitationsThe four defects, and how to fix each

347 system prompts defeat caching · commit 5dff5c4

Pricing

Start free. Pay when it is load-bearing.

Free

For trying it on a side project.

  • 2,500 events per month
  • Live usage dashboard
  • Core cost analytics
  • Anthropic, OpenAI + Gemini tracking
  • Community support
Start free

Pro

For a team shipping one product.

  • Unlimited events
  • Budgets & alerts per agent, team, or org
  • Smart model routing engine
  • Prompt analysis & caching insights
  • Cost forecasting & anomaly detection
  • Priority support

Business

For several teams sharing one bill.

  • Everything in Pro
  • Model cost optimization engine
  • AI optimizer agent
  • Team management & roles
  • Custom integrations
  • Dedicated support

Enterprise

For regulated and high-volume estates.

  • Everything in Business
  • Private deployment options
  • Custom contract and SLA
  • Named support engineer
  • Onboarding and migration help

Built to stay out of your way

One line to adopt. Nothing stored. Nothing blocked if we go down.

1 line
of code to integrate
500ms
from LLM call to live spend
2,500
events tracked free, every month
100%
fail-open, we never block your app
0
prompts or completions ever stored
Capsera
Live spendRealtime
OpenAI
$431
Anthropic
$187
Gemini
$24
Budget used$642 of $3,500
18%
Open the dashboard
reasoning_agent$301
data_extractor$164
support_bot$88

The invoices nobody expected

Every quote below is from a public post by someone running language models in production without a governance layer. We hit the same wall ourselves, which is the reason Capsera exists.

Burning questions about Capsera

Simple answers to make things clear.

Tracking never does: events go into a non-blocking in-memory queue and a background thread flushes them every 500ms. Budget enforcement adds one lightweight pre-call check, and it is optional. Turn it off and Capsera is fully out of your call path. Either way, if the backend is ever unreachable, everything fails open and your calls proceed untouched.

Enforcement lives in the SDK, inside your own process. Before each provider call it checks the budgets that match the calling agent; a budget with a block action stops the call before any provider spend, while throttle and downgrade reshape it. If Capsera is ever unreachable the check fails open, so you lose enforcement for that moment, never availability.

Thirteen, across three kinds. Model APIs: Anthropic, OpenAI, Google, Mistral, Cohere, DeepSeek and Kimi. Inference platforms: Groq, Together, Fireworks and Cerebras. Cloud model services: Vertex AI and Bedrock. Sync and async clients both. The SDK patches the provider libraries automatically at init, so there is no proxy, no gateway, and no change to your existing call sites.

Name each agent where it makes its calls, with a decorator or a context manager, then give it a budget envelope with an action at the limit: alert, throttle, or block. Because identity is inherited through the call context, an agent that spawns sub-agents covers them too, so putting every agent on a budget does not mean enumerating them by hand. Budgets can also be scoped to a team or the whole organisation, and the scopes compose.

Yes, and that is the main reason to use Capsera rather than a provider spend limit. A key-level cap stops every agent sharing that credential when one misbehaves; a per-agent envelope stops only the offender and leaves the rest running. Budgets scope to a single agent, a team, or the organisation, over daily, weekly, or monthly periods with calendar-aware resets.

Capsera instruments the provider clients rather than the framework, so calls made through LangGraph, CrewAI, LangChain, AutoGen, LlamaIndex or plain SDK code are all captured, and helpers attribute spend to LangGraph nodes and CrewAI agents and tasks. One honest caveat: CrewAI routes some calls through litellm internally, so direct client patching can miss those. Framework-level interception that closes this gap is the next scanner change.

Never. Prompt analysis extracts structural metadata only: token estimates, message counts, cache usage, context window utilization. The text of your prompts and completions stays in your infrastructure.

Budgets scope to a single agent, a team, or your whole organization, over daily, weekly, or monthly periods with calendar-aware resets. Each budget picks its action at the limit, either alert, throttle, or block, and you get an alert at your threshold (80% by default) plus an exceeded alert at 100%.

Costs are computed from exact per-model pricing tables using decimal arithmetic, based on the real input, output, and cache token counts returned by each API response, not estimates.

Ingestion returns a clear limit response with your current usage and reset date, and the SDK stops sending events until the month rolls over. Hitting the plan limit never blocks your LLM calls. Only tracking pauses.

Still have questions?

Get Assistance

Where to go next

  • Documentation

    Install the SDK, name your agents, set an envelope. One line to adopt.

  • Who it is for

    Six industries and six roles, and which part of the problem each one meets first.

  • Evidence

    Our own scan of 133 public agent repositories, with the method and the limitations.

  • Guides

    How to evaluate a cost tool, and how to find the agents already in your codebase.

  • Journal

    Field notes, benchmarks and everything else we publish, filterable.

  • About Capsera

    Why this exists, who is building it, and what we will not claim.

Know exactly
where every token goes.

Stop runaway agents before they become runaway invoices.