Want to appear here? Talk with us
TOGETHER WITH KION
Is Your AI Spend Paying Off?
The real challenge with managing AI spend is understanding not only where the money is going, but whether it’s creating business value.
That means being able to connect token spend to the teams, use cases, and business outcomes behind it, then deciding where to set guardrails and where continued investment makes sense.
Our latest webinar recap introduces a practical See It. Control It. Scale It. framework:
✓ See It: Create visibility and accountability for AI token spend
✓ Control It: Put budgets, policies, and guardrails in place
✓ Scale It: Connect spend to outcomes and determine where to invest next
The critical question isn’t how much you’re spending on AI, but what you’re getting for that spend.
See how FinOps teams can evolve their approach to manage AI spend from cost to outcome.
AI COST OPTIMIZATION
Managing Token Spending without Choking Off Opportunity
A lower AI bill is not always a better result. Bain argues that strict spending caps can stop useful work while missing the few tasks that drive large costs.
Use cheaper models for simple work, such as email drafts and document summaries.
Support wider AI use when it saves time or improves results, such as preparing sales proposals.
Set tighter checks for AI agents that can run for long periods without a clear stopping point.
Track what it costs to resolve a customer issue, review a contract, or produce usable code.
Bain says a new AI process may cost 10 times more than it should before teams improve it.
Lower token prices do not guarantee lower bills if usage grows faster. Let business teams own routine AI spending, with central checks for high-risk work.
WEBINAR
Your Cloud Commitments Are Changing. Are They Still Saving You Money?
Join us on October 14th for a live online session exploring how to build a cloud commitment strategy that continuously adapts to changing workloads.
You'll discover:
How to balance savings with flexibility
How rolling purchases can reduce lock-in risk
Why manual commitment optimization doesn't scale
📅 October 14th
⏰ 12:00 PM ET / 5:00 PM UK
📍 Online — Live
AI PROVIDERS
OpenAI cuts API tiers to three, Haiku 5.5 pricing lands, Google adds pay-as-you-go
Anthropic
Claude Haiku 5.5 is the new default Haiku in Claude Code, with 1M context. It costs $0.10/$0.50 per million tokens, and prompts over 100K tokens cost more.
Claude Code v2.1.288 fixes spend tracking. Prompt cache writes and streamed turns are now counted right, so spend numbers are more accurate.
Claude Code v2.1.290 and v2.1.291 add new metadata, like serverToolUses, agentId, and ceiling. This helps you see which agent spent what.
Claude Opus 5.5 and Sonnet 5.5 are now in Gemini Enterprise AI developer tools. They are billed through Cloud Billing and count toward your project spend cap.
Pay-as-you-go above quota now covers more Gemini Enterprise editions, including Frontline, EDU, and Emerging Market. When pooled quota runs out, teams keep working and pay as they go.
OpenAI
API usage tiers dropped from five to three: Build, Launch, and Grow. You move up on your own as your total credit purchases grow, so billing is easier to follow.
A new GPT-6 guide shows how to pick models and tune reasoning effort. These are two big ways to control token spend.
VIDEOS & PODCASTS
How to Measure AI ROI & Control AI Spending
Discover how to manage AI costs, allocate AI spending, set effective budgets, and measure AI ROI. Rick Haggard from nOps explains the AI cost journey, governance, optimization, and how to connect AI spending with real business outcomes.
AI COST ATTRIBUTION
Tokens Are Now the Atomic Unit of AI Spend. Nobody Allocates Atoms
Your AI bill can show what you spent without showing which task caused the cost.
Tania Fedirko argues that FinOps teams need to track each job, not just count tokens or group costs by account.
A “run” is one job, such as handling a customer refund. Several AI agents may work on that job, but their costs should roll up to the same owner.
Her advice:
Put an agent ID and a run ID on every call, and link each child run to its parent.
Check each call’s expected cost against the job’s budget before sending it.
Keep blocked calls as zero-cost records to show where limits stopped spending.
Choose a clear rule for shared costs, and show any costs that still lack an owner.
Match call records to the final bill, then freeze the month’s totals.
FOCUS updates will improve AI billing data, but providers cannot supply your business context.
AI COST MANAGEMENT
AI Spend Budget Busters: Expanding the Budget Planning Lens
Your AI budget can look sound until testing, extra tools, and idle systems push the bill past your plan.
The FinOps Foundation says teams need to budget for the total cost of AI, not just tokens.
Tokenomics estimates that tokens and API fees account for only 10–25% of total AI spend once costs such as hardware, energy, software, and staff are included.
Four steps can help close the gap:
Count the full bill, including testing, monitoring, storage, staff time, and systems waiting to go live.
Review sales plans and product releases early, since each test and customer rollout adds costs.
Track AI tools bought by separate teams to catch repeat purchases and unused spending commitments.
Plan for model changes, aging systems, and duplicate tools after a merger.
Start with clear cost tracking, then move to monthly forecasts and automated spending limits.
🎖️ MENTION OF HONOUR
They Cut 38% of Their Tokens. Their AI Bill Went Up
At Tokenomicon Amsterdam, one team reported using 38% fewer tokens but paying 6.8% more because the model spent longer thinking and took more turns. Track the cost of finished work, not just token counts.
Test the full bill: Include input tokens, output tokens, and answer quality before rolling out a change.
Check default settings: Lundbeck cut one stream of AI spend by more than 60% after finding requests went to a costly region.
Fix each task: Infracost reported 45% savings across its product AI spend by finding issues such as missing caching and repeated work.
Set shared controls: Build caching, batch processing, spending caps, and limits on repeated steps into your tools.
Link AI sessions to code changes and closed tickets so you can track cost per delivered feature.
Save 20% on AI Value & FinOps Certifications
The job market is hungry for certified professionals who can prove results. Don't let your company's budget leak because of a lack of specialization.
Use code: FINOPSWEEKLY_20 to get an instant 20% discount on the most prestigious certification bundles:
FinOps Certified Practitioner
FinOps AI Value
FinOps Technology Value
FinOps Certified Engineer
FinOps Certified FOCUS Analyst
AI Economics Community
Join the Slack community where the real AI Money conversations happen
