Get answers to your AI cost issues:
Pick a path — every link goes straight to the calculator.
Interactive pricing breakdowns with model positioning cards, full rate variants (batch, caching, long context), subscription plans, partner programs, and community-verified gotchas. Verified daily, cross-checked against vendor docs.
See where your AI money actually goes.
Lena Park · Freelance designer paying for 6 AI subscriptions
Upload your ChatGPT, Claude, Cursor, Perplexity, or Midjourney receipt. See where your subscription dollars actually go and get cheaper-tier or shared-plan…
Open CalculatorPriya Patel · Product Manager launching her first AI feature
Walk through the four numbers that drive every LLM bill - input tokens, output tokens, requests/day, and model choice. Live pricing across 17 vendors.
Open CalculatorSarah Chen · VP Engineering at a 50-person SaaS
Estimate monthly burn for coding agents and autonomous workflows across 4 vendors. Walks through Sarah's 5-dev team scenario with live pricing and 3-year…
Open CalculatorDiana Sokolov · CTO at a 250-person Series D
Single-vendor AI is a board-level risk. Quantify your concentration, model migration cost, and design the multi-vendor strategy that won't bankrupt you.
Open CalculatorYara Hassan · VP Operations preparing a board update
Total cost of AI ownership in 5 questions. Inference + ops + tooling + headcount + risk. CFO-ready estimate in 2 minutes.
Open CalculatorSpecific levers to cut your AI bill.
Marcus Lee · Senior Engineer told to 'use the cheap model' for a new feature
Cheapest LLM for your workload - by tier, by task type, by quality threshold. Updated daily as vendors shift pricing. Beyond the per-token table.
Open CalculatorPriya Patel · Eng Lead at a 70-person AI startup
Route easy queries to cheap models, hard queries to premium. Real architecture for cutting AI bills 40-60% without quality drop. Implementation patterns +…
Open CalculatorHiroshi Tanaka · Backend Engineer at a 30-person SaaS
Anthropic charges 10% for cached input tokens. Find the cache-hit rate that makes setup worthwhile - and the workloads where caching saves 30-50%.
Open CalculatorTariq Hassan · Engineering Manager at a 50-person SaaS
Most vendors offer 50% off batch processing. The question isn't 'is batch cheaper' - it's 'what fraction of your workload is actually batch-eligible?'
Open CalculatorCarlos Mendoza · Senior Engineer asked to cut AI bill 30%
Prompt compression, output structure, distillation, smart truncation. Five techniques to cut your AI token bill 30-50% without dropping quality.
Open CalculatorProject AI cost at scale, set budgets, predict overages.
Diana Park · Head of Product at a 30-person Series A SaaS
Most AI bills aren't linear at scale. Find the cliffs - rate limits, tier jumps, latency walls - before they find you. Live pricing, real benchmarks, vendor…
Open CalculatorRobert Tanaka · FinOps lead at a 200-person SaaS
Project your AI bill month-by-month for 12 months. Surface budget breaches before they happen. Models growth + seasonality + vendor pricing trends.
Open CalculatorDaniel Liu · VP Product at a 100-person SaaS
Split your annual AI budget across product features by ROI priority. Avoid overspending on shiny features at the cost of high-ROI utility ones.
Open CalculatorCarlos Mendez · Engineering manager owning the AI cost line
Project when your AI spend hits the budget cap. Models trend + variance + vendor pricing. Get the breach date and the optimization runway you have left.
Open CalculatorNaomi Bell · Pricing Strategy Lead at a Series B SaaS
Revenue per AI request vs cost per AI request. Find break-even, gross margin, and the price you can charge. CFO-defensible math for AI feature pricing.
Open CalculatorSee where your AI subscription money goes, cut tokens, plan a dev stack at low cost.
Lena Park · Freelance designer paying for 6 AI subscriptions
Upload your ChatGPT, Claude, Cursor, Perplexity, or Midjourney receipt. See where your subscription dollars actually go and get cheaper-tier or shared-plan…
Open CalculatorSara Patel · Freelance designer + part-time tutor
Picking your AI subscription? Real comparison of ChatGPT Plus, Claude Pro, Gemini Advanced, Cursor Pro. By use case, by quality, by price.
Open CalculatorDaniela Costa · Recent grad, freelancing on a tight budget
Free AI tiers are real. ChatGPT, Claude, Gemini all give meaningful free use. Find which combination covers your needs without spending a dollar.
Open CalculatorMarcus Wei · Senior backend developer + side projects
Best developer AI stack at $20-100/mo. Cursor, GitHub Copilot, Claude Code, ChatGPT, Codeium. Real comparison + when to combine.
Open CalculatorAgent loops, multi-step workflows, voice agents, full stacks.
Sarah Chen · VP Engineering at a 50-person SaaS
Estimate monthly burn for coding agents and autonomous workflows across 4 vendors. Walks through Sarah's 5-dev team scenario with live pricing and 3-year…
Open CalculatorAisha Patel · Staff Engineer building a multi-step research agent
Multi-turn agents (ReAct, AutoGPT, function-calling) have compounding token costs. Model the per-task cost + runaway risk before deploying. Live pricing…
Open CalculatorQuincy Ross · Tech Lead architecting an internal agent platform
Agents aren't one cost - they're five. LLM + tool calls + memory + orchestration + observability. Real architecture math for production agent systems.
Open CalculatorAiyana Crow · Tech Lead at a voice-first customer service startup
Voice agents combine STT + LLM + tools + memory + TTS or voice-native models. Real architecture math for production voice products.
Open CalculatorEsme Vasquez · ML Engineer building a video-and-document Q&A product
Multimodal RAG combines image embeddings, audio transcription, and text retrieval. Real architecture math for production multimodal apps.
Open CalculatorPipeline, embeddings, chunking, hybrid search.
Krishna Iyer · Tech Lead designing a customer-facing knowledge bot
RAG isn't one cost - it's five. Embedding indexing + storage + query embedding + retrieval + LLM read. Real architecture math for production RAG.
Open CalculatorMaya Iyer · ML Lead at a 80-person FinTech
RAG ships fast and adapts to fresh data; fine-tuning is cheaper at scale. Find the break-even - and avoid choosing wrong on a 6-month commitment.
Open CalculatorOlivia Garrett · Solutions Engineer building a knowledge base RAG
Embeddings are 10-30× cheaper than chat - but volume adds up. Index cost + query cost + re-embedding triggers. Real RAG pipeline math.
Open CalculatorTomoko Sato · ML Engineer iterating on a RAG retrieval system
Chunk size is the most-tweaked, least-understood RAG parameter. Find the size that maximizes recall while controlling cost - workload-specific.
Open CalculatorAndre Williams · Senior Engineer building product search
Hybrid retrieval (BM25 + dense) beats pure semantic for most workloads. Real cost vs recall math, and when the extra complexity pays back.
Open CalculatorPricing across 25 LLMs, 9 embedding models, 7 vision models, 12 audio services. Take a number to your CFO that holds up to scrutiny.
Total cost of ownership across 6 pillars. Workload × vertical × cloud aware. Combines our 38 calcs where precision matters with industry-typical ranges (cited) where it doesn't. Tool handoffs to ToolsInfo so YOU pick the vendor.
3 dropdowns + 4 numbers → full TCO + ROI + payback in ~90 sec.
Open Calculator7-step wizard with calc handoffs & ToolsInfo drill-downs. 5-15 min for procurement-grade precision.
Open CalculatorPre-filled for MSPs running AI workloads on AWS. Multi-tenant security + per-client compliance built in.
Open CalculatorOn June 15, 2026, Anthropic's Agent SDK credit goes live. Every Pro / Max / Team Premium / Enterprise Premium plan starts including a separate monthly SDK credit equal to the plan's base price. Pro at $20/mo effectively becomes $40 of value. GitHub Copilot transitions to usage-based billing two weeks earlier on June 1. Your team's subscription math just changed. Two new calculators answer the question every builder and FP&A director will be asked this quarter.
Should you switch from Claude API to Pro/Max+SDK? Net savings, payback, ranked across 6 Anthropic plans. Includes Team Standard 'NOT eligible' trap detection.
Open Calculator4-step wizard across Anthropic, OpenAI, Google, GitHub, Microsoft. Persona-driven (Eng leader / Solo founder / FP&A / Consultant). 13 plans ranked by your…
Open CalculatorGet the full integrated cost on one screen. Each stack composes the right atomic calcs (planner + executor + verifier; ingest + storage + queries; STT + LLM + TTS) and surfaces what-if savings. Playbooks walk you through step by step.
Quincy Ross · Tech Lead architecting an internal agent platform
Agents aren't one cost - they're five. LLM + tool calls + memory + orchestration + observability. Real architecture math for production agent systems.
Open CalculatorEsme Vasquez · ML Engineer building a video-and-document Q&A product
Multimodal RAG combines image embeddings, audio transcription, and text retrieval. Real architecture math for production multimodal apps.
Open CalculatorAiyana Crow · Tech Lead at a voice-first customer service startup
Voice agents combine STT + LLM + tools + memory + TTS or voice-native models. Real architecture math for production voice products.
Open Calculator4-step guided journey · models, routing, caching, loop overhead. Resume anytime.
Open Calculator5-step guided journey · ingest, vector DB, retrieval, reranker, LLM read.
Open CalculatorPriya Patel · Product Manager launching her first AI feature
Walk through the four numbers that drive every LLM bill - input tokens, output tokens, requests/day, and model choice. Live pricing across 17 vendors.
Open CalculatorSide-by-side comparison of every model by price, context, modality.
Open CalculatorMarcus Lee · Senior Engineer told to 'use the cheap model' for a new feature
Cheapest LLM for your workload - by tier, by task type, by quality threshold. Updated daily as vendors shift pricing. Beyond the per-token table.
Open CalculatorJames Wong · Senior Engineer building a customer support assistant
Paste your real prompt, get accurate token count + monthly cost projection across 17 vendors. Stop guessing at token counts that swing your bill 3-5×.
Open CalculatorHannah Park · Senior Engineer at a doc-analysis startup
1M-token context windows enable new use cases - and double your bill. Find the threshold where chunking + RAG beats long-context, and where it doesn't.
Open CalculatorPriya Patel · Eng Lead at a 70-person AI startup
Route easy queries to cheap models, hard queries to premium. Real architecture for cutting AI bills 40-60% without quality drop. Implementation patterns +…
Open CalculatorHiroshi Tanaka · Backend Engineer at a 30-person SaaS
Anthropic charges 10% for cached input tokens. Find the cache-hit rate that makes setup worthwhile - and the workloads where caching saves 30-50%.
Open CalculatorTariq Hassan · Engineering Manager at a 50-person SaaS
Most vendors offer 50% off batch processing. The question isn't 'is batch cheaper' - it's 'what fraction of your workload is actually batch-eligible?'
Open CalculatorCarlos Mendoza · Senior Engineer asked to cut AI bill 30%
Prompt compression, output structure, distillation, smart truncation. Five techniques to cut your AI token bill 30-50% without dropping quality.
Open CalculatorAditi Sharma · VP Engineering deciding on AI sales coaching tooling
Should you buy a vertical AI SaaS (Cresta, Glean, Harvey) or build your own with OpenAI/Anthropic APIs? Real cost math + non-cost factors + decision framework.
Open CalculatorMarcus Lee · CFO at a 250-person professional services firm
MIT NANDA reports 95% of GenAI pilots fail to show ROI. This calculator + guide pricing the workload - hours saved, revenue lift, risk avoided, AI spend -…
Open CalculatorNaomi Bell · Pricing Strategy Lead at a Series B SaaS
Revenue per AI request vs cost per AI request. Find break-even, gross margin, and the price you can charge. CFO-defensible math for AI feature pricing.
Open CalculatorDaniel Liu · VP Product at a 100-person SaaS
Split your annual AI budget across product features by ROI priority. Avoid overspending on shiny features at the cost of high-ROI utility ones.
Open CalculatorRobert Tanaka · FinOps lead at a 200-person SaaS
Project your AI bill month-by-month for 12 months. Surface budget breaches before they happen. Models growth + seasonality + vendor pricing trends.
Open CalculatorLila Reyes · Operations Director at a 60-person agency
Annual AI commitments save 10-20% but lock you in for 12 months. Find the conditions where it pays - and the ones where flexibility matters more.
Open CalculatorDiana Park · Head of Product at a 30-person Series A SaaS
Most AI bills aren't linear at scale. Find the cliffs - rate limits, tier jumps, latency walls - before they find you. Live pricing, real benchmarks, vendor…
Open CalculatorHannah Schmidt · Procurement Manager at a 400-person German enterprise
AI vendor pricing is USD-denominated. Convert to your local currency for budgeting, finance, and procurement. Live FX rates, current vendor pricing.
Open CalculatorCarlos Mendez · Engineering manager owning the AI cost line
Project when your AI spend hits the budget cap. Models trend + variance + vendor pricing. Get the breach date and the optimization runway you have left.
Open CalculatorHannah Kim · Senior FinOps Manager presenting to CFO quarterly
Model your AI spend across quarters with seasonality, growth rate, and pricing assumption variance. Brief CFO with confidence intervals, not point estimates.
Open CalculatorOlivia Garrett · Solutions Engineer building a knowledge base RAG
Embeddings are 10-30× cheaper than chat - but volume adds up. Index cost + query cost + re-embedding triggers. Real RAG pipeline math.
Open CalculatorAisha Patel · Staff Engineer building a multi-step research agent
Multi-turn agents (ReAct, AutoGPT, function-calling) have compounding token costs. Model the per-task cost + runaway risk before deploying. Live pricing…
Open CalculatorSarah Chen · VP Engineering at a 50-person SaaS
Estimate monthly burn for coding agents and autonomous workflows across 4 vendors. Walks through Sarah's 5-dev team scenario with live pricing and 3-year…
Open CalculatorFaisal Ahmad · ML Engineer at a 100-person legal tech company
Fine-tuning math: training compute + tokens + base model selection + inference savings. When custom models pay back - and when they don't.
Open CalculatorMaya Iyer · ML Lead at a 80-person FinTech
RAG ships fast and adapts to fresh data; fine-tuning is cheaper at scale. Find the break-even - and avoid choosing wrong on a 6-month commitment.
Open CalculatorVihaan Reddy · Backend Engineer choosing vector storage
Vector DB pricing varies 10× between hosted SaaS and self-hosted. Storage cost + query cost + ops overhead. Real math for RAG production.
Open CalculatorMei Lin · Product Engineer launching a receipt-OCR feature
Vision pricing is weirder than text. Tile-based, resolution-tier, per-image and per-token mixed. Real math across GPT-5.5 Vision, Claude, Gemini for production.
Open CalculatorSven Mikkelsen · Product Lead at a 40-person customer service tool
Speech-to-text per minute, text-to-speech per character, voice agent stack cost. Whisper, Deepgram, ElevenLabs, OpenAI Realtime - when each wins.
Open CalculatorDiana Sokolov · CTO at a 250-person Series D
Single-vendor AI is a board-level risk. Quantify your concentration, model migration cost, and design the multi-vendor strategy that won't bankrupt you.
Open CalculatorWei Chen · VP Engineering at a 200-person Series C startup
Self-hosting Llama 3 / Mistral on GPUs vs API: where break-even hits. Includes ops cost, capacity utilization, and the privacy multiplier.
Open CalculatorKrishna Iyer · Tech Lead designing a customer-facing knowledge bot
RAG isn't one cost - it's five. Embedding indexing + storage + query embedding + retrieval + LLM read. Real architecture math for production RAG.
Open CalculatorTomoko Sato · ML Engineer iterating on a RAG retrieval system
Chunk size is the most-tweaked, least-understood RAG parameter. Find the size that maximizes recall while controlling cost - workload-specific.
Open CalculatorAndre Williams · Senior Engineer building product search
Hybrid retrieval (BM25 + dense) beats pure semantic for most workloads. Real cost vs recall math, and when the extra complexity pays back.
Open CalculatorSara Patel · Freelance designer + part-time tutor
Picking your AI subscription? Real comparison of ChatGPT Plus, Claude Pro, Gemini Advanced, Cursor Pro. By use case, by quality, by price.
Open CalculatorDaniela Costa · Recent grad, freelancing on a tight budget
Free AI tiers are real. ChatGPT, Claude, Gemini all give meaningful free use. Find which combination covers your needs without spending a dollar.
Open CalculatorMarcus Wei · Senior backend developer + side projects
Best developer AI stack at $20-100/mo. Cursor, GitHub Copilot, Claude Code, ChatGPT, Codeium. Real comparison + when to combine.
Open CalculatorYuki Watanabe · Solo content creator (YouTube + TikTok)
Solo creators and content makers - pick the right AI tool stack. Image, music, voice, video, writing. Real cost vs ROI math for monetizing creators.
Open CalculatorJamal Brooks · Parent of teenagers + small business owner
Sharing AI subscriptions across family or small team? ChatGPT Team, Claude Team, Gemini family options. Real per-user math + when family plan beats individual.
Open CalculatorUS vs EU vs APAC AI region pricing. Data residency trade-offs across Bedrock, Azure OpenAI, Vertex.
Open CalculatorInstant routing to the right product line, the exact playbook, and the tools that match your problem.
👋 Tell me what’s going on. I’ll surface the right frameworks, tools, and playbooks, plus which product line fits.
Pick the problem closest to yours:
Your AI bill isn't just tokens. It shows up across six dimensions — most of which never appear on a dashboard.
When enterprises track “AI cost,” they count tokens, GPU hours, and API calls. But the true cost extends far beyond the invoice.
These are the dimensions of the invisible bill.
The categories below are pulled from our sister site ToolsInfo.com — 115K+ workflow tools that pair with AI cost optimization.
from sister website ToolsInfo.com →
Every AI feature has a per-request compute cost. Scale from 1K to 1M users and your bill can 10x while revenue 2x. This is where traditional FinOps playbooks break — they were written before inference became the biggest line item.
Scale your cloud footprint without exceeding your budget.
Boost cloud savings by automating the lifecycle of your instances
Scale cloud workloads while cutting compute costs by up to 90%
Scale workloads affordably with automated spot market bidding
Manage cloud budgets to eliminate waste and increase margins
Close the gap on cloud overspending with automated savings
Boost your cloud ROI by eliminating waste and hidden costs
Boost cloud savings by automating long-term discount strategies
Manage AI expenses and maximize ROI by eliminating wasted compute
Manage AI costs with granular visibility and governance tools.
Track profitability by linking cloud costs to business value.
Boost compute capacity while cutting costs using preemptible VMs.
Automate spot bidding to maximize savings and ensure app stability.
Scale savings by predicting and securing future capacity needs.
Track LLM expenses to slash your monthly API bills immediately
Track cloud spending to eliminate waste and maximize your budget.
Scale your cloud ROI by eliminating wasted reserved capacity
Manage spot fleets to cut compute costs without sacrificing uptime.
Track cloud spending by department to drive financial accountability.
Automate RI lifecycle management to slash cloud infrastructure costs.
Streamline RI liquidity to recover capital from unused commitments
Automate workload migration to slash cloud costs without downtime
Track cloud spend to eliminate waste and maximize your ROI
Scale cloud resources efficiently while automatically reducing monthly spend
from sister website ToolsInfo.com →
Hallucinations cost enterprises $14,200 per employee annually in human-in-the-loop verification time. Bias drift in regulated industries triggers fines in the billions. Every AI output that reaches a customer is a liability event waiting to happen.
Streamline model launches with automated accuracy scoring
Manage AI performance by choosing the best model for your needs
Automate LLM testing to ensure consistent output quality at scale
Automate AI quality checks to ship production-ready models faster
Boost AI accuracy and prevent hallucinations in your production
from sister website ToolsInfo.com →
Every regulated industry (healthcare, finance, legal, government) is rebuilding AI programs to stay compliant. Miss an audit trail, skip a model card, fail to document bias — and you're not just losing money, you're losing the right to operate.
Streamline regulatory audits with automated AI governance
Automate regulatory compliance to reduce legal risk across teams
from sister website ToolsInfo.com →
The fastest way to lose 10x what you saved on AI infrastructure is a data leak through a prompt. Clean rooms, PII detection, red-team testing, and prompt injection defense aren't optional — they're the difference between AI as competitive advantage and AI as breach vector.
Manage compliance risks and protect sensitive data automatically
Track model vulnerabilities to prevent breaches and ensure safety
Scale data sharing safely without compromising user privacy
Manage LLM security risks to ensure safe and compliant AI
Scale data usage safely while protecting sensitive user privacy
Manage AI security risks by identifying vulnerabilities early
Manage AI security risks and prevent model data leaks effortlessly
Close security gaps by identifying model flaws before launch
Automate LLM security to protect your data and brand reputation
from sister website ToolsInfo.com →
A single domain fine-tune run costs $8K-$60K in compute alone, plus a $200K/year MLE to supervise. Moving a model from notebook to production requires MLOps infrastructure most teams underestimate by 3-5x. This is where R&D budgets die.
Manage the entire ML lifecycle to ship reliable models faster
Manage the full ML lifecycle to ensure reliable model performance
Automate end-to-end ML lifecycles to accelerate model delivery
Streamline model deployment with consistent, real-time ML data
from sister website ToolsInfo.com →
A single infinite-loop agent can rack up $20K overnight. Silent model drift reduces quality for months before anyone notices. Without observability purpose-built for AI workloads, you're flying blind — and the bill is how you learn.
Automate complex workflows to save hours of manual effort
Track model performance to prevent costly production errors
Build better AI responses with systematic prompt optimization
Automate complex workflows with custom AI that scales with you
Automate incident response to prevent outages before they happen
Route queries across providers ? failover, cost optimization, rate-limit handling.
Track request paths to resolve microservice bottlenecks faster
Track resource health to prevent downtime and reduce cloud costs
Streamline incident response by identifying hidden performance gaps
Automate daily scheduling to reclaim your productive hours
Build proprietary models that provide superior accuracy
Boost system uptime by predicting and resolving IT issues faster.
Build intelligent agents to resolve customer issues instantly
Automate complex processes to reclaim hours of manual work
Manage AI model accuracy to deliver reliable user experiences.
Build human-like voice experiences to increase user engagement.
Boost AI accuracy with real-time data to eliminate hallucinations.
Streamline software delivery and ship complex features faster.
Manage security risks by turning raw logs into actionable data
Streamline troubleshooting to keep your systems running smoothly.
Boost decision-making by analyzing video, text, and audio data
Build intelligent features faster without training custom models
Automate incident response to maintain 24/7 system availability
Boost application uptime by identifying server-side bottlenecks
Streamline system observability with vendor-neutral data.
Track service connections to prevent outages during system changes.
Manage LLM risks and ensure brand safety in every interaction
Streamline troubleshooting with real-time system visibility
Boost conversion rates by optimizing real-world site performance
Track distributed requests to find and fix performance bottlenecks
Boost application performance by identifying slow user transactions
Track system logs in real-time to resolve production issues faster.
Streamline troubleshooting to reduce mean time to recovery
Build autonomous workflows that execute complex business tasks
Scale AI performance with high-speed hosting and low latency
Track request paths to pinpoint bottlenecks in complex microservices
Boost application performance and reduce MTTR with total visibility.
Close visibility gaps by aggregating all logs into a single view
Streamline cloud traffic to eliminate latency and ensure connectivity
Automate metrics collection to make proactive scaling decisions
Automate customer support to resolve inquiries instantly at scale.
Scale infrastructure confidently while ensuring maximum system uptime
Boost system uptime by preventing critical IT failures before they happen
Manage logs effortlessly across dynamic, ephemeral cloud systems
Track resource utilization to optimize performance and cut costs.
Track server health to prevent downtime before it impacts customers.
Boost application uptime and deliver a seamless user experience.
Track system health in real-time to ensure maximum service reliability.
Automate repetitive web tasks to save hours of manual work.
Weekly-refreshed AI cost intelligence by persona/vendor, 2.5 years of pricing history (charted), and one-click cloud cost connectors for AWS, Azure, and GCP.
15 vendors × 5 audience lenses = 75 persona-led guides · cross-vendor synthesis · procurement intel competitors don't surface
Daily snapshots of every major AI vendor's pricing since Sept 2023. 62,000+ historical price points across OpenAI, Anthropic, Google, Mistral — chartable, citable, CSV-exportable. The only public AI pricing archive with this depth.
Stop guessing — connect your real cloud bills. Read-only IAM role (you create it via CloudFormation Quick-Create — no credentials shared). We pull Cost Explorer + Bedrock data daily, surface savings recommendations, feed real numbers into every calculator on this site.
Three productized brands. Three sizes each. Personal · SMB · Enterprise.
Where your AI money actually goes.
You get a clear, honest diagnosis of your current AI spend.
What to change. With dollar savings.
Diagnosis + 3 specific optimization levers. Plus ToolsInfo workflow pairing.
What it'll cost at scale.
Multi-scenario TCO modeling — inference + infra + MLOps + risk.
If you still need help, we're here for $19.99 to enterprise.
Last-verified date is the most recent successful daily snapshot
(aicost_pricing_snapshots) or, when no snapshot exists yet,
the latest successful crawler run (aicost_crawler_runs).
10 of 10
vendors are currently verified. Aggregator services (TokenCost, AI Pricing Guru, etc.)
are not listed.
Derived from industry conventions, not directly published by the vendor. Typical conventions: cached input = 10% of base (90% off), Batch API = 50% of base (50% off).
| Vendor / Model | Field | Why it’s inferred |
|---|---|---|
| Anthropic — Claude Sonnet 4.6 | cachedInput |
Derived at 10% of input rate — Anthropic publishes 90% cache-hit discount on this tier. |
| Anthropic — Claude Sonnet 4.5 | cachedInput |
Derived at 10% of input rate; same 90% cache-hit convention as Sonnet 4.6. |
| Anthropic — Claude Sonnet 4.5 | batchInput |
Derived at 50% of standard input — Anthropic documents uniform 50% Batch discount. |
| Anthropic — Claude Sonnet 4.5 | batchOutput |
Derived at 50% of standard output — Anthropic documents uniform 50% Batch discount. |
| Anthropic — Claude Haiku 4.5 | cachedInput |
Derived at 10% of input rate — Anthropic 90% cache-hit discount convention. |
| OpenAI — GPT-5.4 Mini | cachedInput |
Derived at 10% of input — OpenAI documents automatic 90% discount on cache hits across GPT-5.x tier. |
| OpenAI — GPT-5.4 Nano | cachedInput |
Derived at 10% of input — OpenAI 90% cache-hit convention. |
| OpenAI — GPT-5.4 Nano | batchInput |
Derived at 50% of input — OpenAI Batch API uniform 50% discount. |
| OpenAI — GPT-5.4 Nano | batchOutput |
Derived at 50% of output — OpenAI Batch API uniform 50% discount. |
| OpenAI — GPT-5.4 Pro | cachedInput |
Derived at 10% of input — OpenAI 90% cache-hit convention. |
| OpenAI — GPT-5.4 Pro | batchInput |
Derived at 50% of input — OpenAI Batch API uniform 50% discount. |
| OpenAI — GPT-5.4 Pro | batchOutput |
Derived at 50% of output — OpenAI Batch API uniform 50% discount. |
| OpenAI — GPT-5.2 | cachedInput |
Derived at 10% of input; no residency uplift. |
| OpenAI — GPT-5.2 | batchInput |
Derived at 50% of input. |
| OpenAI — GPT-5.2 | batchOutput |
Derived at 50% of output. |
| OpenAI — GPT-5 | cachedInput |
Derived at 10% of input. |
| OpenAI — GPT-5 | batchInput |
Derived at 50% of input. |
| OpenAI — GPT-5 | batchOutput |
Derived at 50% of output. |
| OpenAI — GPT-5.5 Pro | cachedInput |
Derived at 10% of input — OpenAI does not publish a cached rate for *-pro models; using the family convention. |
| OpenAI — GPT-5.5 Pro | batchInput |
Derived at 50% of input. |
| OpenAI — GPT-5.5 Pro | batchOutput |
Derived at 50% of output. |
| OpenAI — GPT-5.2 Pro | cachedInput |
Derived at 10% of input — pro-tier convention. |
| OpenAI — GPT-5.2 Pro | batchInput |
Derived at 50% of input. |
| OpenAI — GPT-5.2 Pro | batchOutput |
Derived at 50% of output. |
| OpenAI — GPT-5.1 | batchInput |
Derived at 50% of input. |
| OpenAI — GPT-5.1 | batchOutput |
Derived at 50% of output. |
| OpenAI — GPT-5 Pro | batchInput |
Derived at 50% of input. |
| OpenAI — GPT-5 Pro | batchOutput |
Derived at 50% of output. |
| OpenAI — GPT-5 Nano | cachedInput |
Derived at 10% of input. |
| OpenAI — GPT-5 Nano | batchInput |
Derived at 50% of input. |
| OpenAI — GPT-5 Nano | batchOutput |
Derived at 50% of output. |
| Google — Gemini 3 Flash | cachedInput |
Derived at 10% of input — Google caching discount convention ~90%. |
| Google — Gemini 3.1 Flash-Lite | cachedInput |
Derived at 10% of input — Google caching convention. |
| Google — Gemini 3.1 Flash-Lite | batchInput |
Derived at 50% of input — Google Batch API uniform 50% discount. |
| Google — Gemini 3.1 Flash-Lite | batchOutput |
Derived at 50% of output — Google Batch API uniform 50% discount. |
| Google — Gemini 2.5 Pro | cachedInput |
Derived at 10% of input. |
| Google — Gemini 2.5 Flash | cachedInput |
Derived at 10% of input. |
| Google — Gemini 2.5 Flash-Lite | cachedInput |
Derived at 10% of input — Google caching convention. |
| Google — Gemini 2.5 Flash-Lite | batchInput |
Derived at 50% of input — Google Batch API uniform 50% discount. |
| Google — Gemini 2.5 Flash-Lite | batchOutput |
Derived at 50% of output — Google Batch API uniform 50% discount. |
| Google — Gemini 2.0 Flash | cachedInput |
Derived at 25% of input per Google 2.0 family caching rates. |
| Google — Gemini 2.0 Flash | batchInput |
Derived at 50% of input — Google Batch API uniform 50% discount. |
| Google — Gemini 2.0 Flash | batchOutput |
Derived at 50% of output — Google Batch API uniform 50% discount. |
| Google — Gemini 2.0 Flash-Lite | cachedInput |
Derived at 10% of input — Google caching convention. |
| Google — Gemini 2.0 Flash-Lite | batchInput |
Derived at 50% of input — Google Batch API uniform 50% discount. |
| Google — Gemini 2.0 Flash-Lite | batchOutput |
Derived at 50% of output — Google Batch API uniform 50% discount. |
| xAI — Grok 4 (legacy) | cachedInput |
Extrapolated at 25% of base. |
Pricing is cross-verified against the
LiteLLM community registry
when available. Daily snapshots are kept in aicost_pricing_snapshots;
every change is logged to aicost_price_changelog with old & new
values for full audit trail. Read the full methodology →
