Start free
ClaudeOpenAIGeminiAnthropic SDKPythonREST APIMCPVS CodeClaudeOpenAIGeminiAnthropic SDKPythonREST APIMCPVS Code
Anthropic SDKPythonREST APIMCPVS CodeClaudeOpenAIGeminiAnthropic SDKPythonREST APIMCPVS CodeClaudeOpenAIGemini

We cut your AI costs. You pay us from the savings.

Tokeven tracks spend across Claude, ChatGPT, and Gemini, flags waste and recommends cheaper alternatives. We take 10% of what you actually save.

2-min installPrompt content never reaches TokevenAccept or ignore recommendationsFree tier forever

Gainshare Pricing

We only earn when you actually save

Gainshare pricing means we only earn when you save real money - calculated against live token prices, not a frozen baseline. You keep 90% of every dollar saved. If we save you nothing, you pay nothing.

Free

$0
per org / month
  • Full cost visibility dashboard
  • Pre-call cost estimates via MCP
  • Model recommendations
  • Local-only privacy (no keys, no prompts)
  • 1,000 tracked calls/month
Start free

Team

10%
of verified savings
  • Everything in Free
  • Unlimited tracked calls
  • Org-wide savings dashboard
  • Per-person savings tracking
  • Savings recommendations
  • Prompt revision suggestions
  • Weekly analysis digest
  • Cross-provider analytics
  • Embeddable widgets
  • CSV, JSON & API export
  • n8n & webhook integrations
Start free

Enterprise

Custom
capped or high-touch terms
  • Everything in Team
  • Dedicated account manager
  • Custom integrations
  • SLA & priority support
  • SSO & SAML
  • Custom data retention
Talk to us

Savings measured against real-time token costs per model. No frozen baselines, no hidden math.

The math on a $50k bill

16%
lower monthly bill
$8,000
saved per month
$7,200
you keep — we take 10% of what we save you
2min
to install and start tracking

How this is calculated — a worked model, not a customer average

A team spending $50,000/mo moves 40% of its Claude Opus 4.8 traffic to Claude Sonnet 4.6. At today’s published token prices that traffic costs 60% of what it did (blended 3:1 input/output), so $20,000 becomes $12,000 — $8,000 saved. Tokeven takes 10% of that ($800); you keep the rest. Your own numbers will differ — that is the point of measuring them.

How it works

From install to real savings in minutes

Two lines of code, zero behavior change. Before each call, your team sees a cost forecast and model alternatives in their terminal. After each call, usage is logged. Nothing else changes. See the full walkthrough
Step 1

Install the wrapper

Two lines of Python wrap your existing Claude, OpenAI, or Google client. Tokeven captures token counts, model, latency, and cost while your calls run exactly as before. Your keys stay local and prompt content never reaches Tokeven.

  • Works with Claude, GPT, and Gemini
  • Runs locally - we never see your prompts
  • Zero config to start
settings.json
1// Claude Code → settings.json
2// pip install "tokeven[mcp]"
3{
4 "mcpServers": {
5 "tokeven": {
6 "command": "tokeven-mcp",
7 "env": { "TOKEVEN_PROJECT": "my-app" }
8 }
9 }
10}
Your Weekly Spend Digest
· tokeven
Mon 9:00 AM
3 optimizations found - save $1,520/mo
Total calls
312
Total cost
$1,847
Avg cost/call
$5.92
Cache hit rate
34%
42% of Opus calls could use Sonnet
$890/mo

You made 63 calls to claude-opus-4-8 with avg prompt length under 200 chars. Sonnet handles these at 40% lower cost with comparable quality.

Enable cache_control on your system prompt
$520/mo

Your 4,200-token system prompt repeats 8x/day across sessions. Adding cache_control would cut 90% off cached input tokens.

Trim repeated context in 3 workflows
$110/mo

3 workflows re-send full conversation history on every call. Summarizing prior turns before re-sending would cut input tokens by ~60%.

View full report in dashboard
Step 2

See cost and alternatives before every call

Before each prompt is sent, Tokeven shows the cost on the current model, what it would cost on alternatives across Claude, GPT, and Gemini, and a confidence score for each. Your team picks the right model in the moment — no policy to set, no black-box rerouting. A weekly digest summarizes the savings and flags patterns.

  • Price forecast shown before every call
  • Model confidence scores across all providers
  • Weekly digest with dollar amounts per recommendation
Step 3

Prove every dollar saved

The savings proof dashboard traces every dollar back to the action that caused it - accepted routing recommendations, cache improvements, prompt optimizations. Per-person, per-team, fully auditable. Savings are calculated against current token prices, so the numbers stay honest as costs change.

  • Only accepted recommendations count
  • Per-person and per-team breakdowns
  • Savings measured against live token prices
Savings Proof, 30 daysExample
Verified Savings
$1,240
Spend Reduction
30%
Weekly Cost Trend
W1W2W3W4W5W6
Cross-Provider Routing
$520
Model Downgrades
$380
Cache Hits
$340

Pre-send cost forecast

Before you send, you see the price

Per-call cost tracking, model breakdowns, caching analytics, and team spend across Claude, GPT, and Gemini - all from metadata your wrapper collects locally. No API keys shared, no prompts stored.
Cost OverviewExample dashboard
Total Cost
$4,200
Saved
$1,240
Total Calls
2,660
Cache Rate
34%
Optimization Insights

GPT 5.5 Pro used for tasks where Gemini Flash matches quality

Route summarization to Flash - 85% cheaper

Save $720/wk

68% of Opus calls are single-turn Q&A

Switch to Sonnet or Gemini Flash for simple lookups

Save $580/wk

System prompts repeat across 89% of sessions (all providers)

Enable prompt caching where supported

Save $320/wk

app.tokeven.com/dashboard/cost
Cost Overview
7 days
30 days
90 days
Total Cost (7d)
$4,200
+12% vs prev
Avg Cost / Call
$0.14
1,847 calls
Cache Hit Rate
34%
↑ from 28%
Daily Cost
MonTueWedThuFriSatSun
Cost by Model
claude-opus-4-8$1,840
gpt-5.6-sol$1,520
claude-sonnet-4-6$920
gemini-3.1-pro$780
gpt-5.5$540
gemini-3.5-flash$220
claude-haiku-4-5$180
Most Expensive Calls
gpt-5.6-sol
$0.48
claude-opus-4-8
$0.42
gemini-3.1-pro
$0.34
claude-sonnet-4-6
$0.28

What changes with Tokeven

Cost awareness
Bill arrives at month end
Every call priced before it's sent
Model selection
Developers guess which model and provider fits
Cost and confidence shown per model before each call — across Claude, GPT, and Gemini
Prompt caching
Repeated system prompts billed at full price
Cacheable patterns flagged, 90% savings on repeats
Team visibility
No idea who spends what
Per-person cost, calls, and savings - fully auditable
Savings proof
"We think we saved money"
Every dollar traced to a recommendation you accepted
Privacy
Prompts sit on vendor servers
Prompt content never reaches Tokeven, metadata only

Interactive Calculator

Run your own numbers

Plug in your current AI spend and see what better model selection, caching, and prompt efficiency would save.

Calculate your savings

$50,000
$5K$500K
25
5500
Assumptions
Average savings rate30%
Tokeven fee (gainshare)10% of savings

Your projected savings

Annual net savings
$162,000
after Tokeven fee
Monthly savings (gross)
$15,000
Tokeven fee (10%)
$1,500
You keep monthly$13,500
That's $540 saved per seat per month
Start saving today

Questions buyers always ask

Your team spends less
when they see the cost upfront

A price forecast before every call. Accepted suggestions reduce your bill. We take 10% of what you actually save.

Start free
Free tier foreverCard-free signupTracking from minute one