01
Blended cost depends on your input:output ratio
Output usually costs 3-4x input. Two models with the same input rate can differ 2x on a real workload, so your token split matters more than the headline price.
LLM API pricing · prompt-cache economics · cost verification
What LLM APIs actually cost.
Real LLM API prices, in USD and CHF. Every number is sourced, and when a rate comes from billing data instead of a pricing page, we say so. Mostly measured on Microsoft Foundry (formerly Azure AI Foundry), plus direct APIs to compare against.
97%
Cost cut from caching alone, measured
$12.42→$1.74
Same task, cache off vs on (CHF 10 → 1.40)
4.5×
Widest cloud resale markup vs direct
7
Providers, all in USD + CHF
Reference
Providers
One page each: the full table in USD and CHF, plus the quirks a pricing page hides, like cache meters, Responses-API-only models, and deployment-tier premiums.
KimiMoonshot AI
Kimi K3 is a 1M-context flagship, now on Microsoft Foundry as well as Moonshot's own API; earlier K2 models are resold on Foundry across tiers.
from $0.60 /M in→
DeepSeek
1M-token context, cheap direct pricing, and Microsoft Foundry resale markups from ~10% to a reported 4.5x — with one Data Zone lane that undercuts Global.
from $0.15 /M in→
GLMZhipu / Z.ai
GLM-5.3-Flash adds a multimodal, half-price direct lane, while GLM-5.3, 5.1 and 5.2 remain aligned on the full-rate tier.
from $0.075 /M in→
OpenAI / Azure OpenAI
Azure normally resells OpenAI 1:1, but a promotional cut on the GPT-5.6 Sol flagship has opened a fresh gap. Deployment type and Responses-API-only variants are the other catches.
from $0.20 /M in→
ClaudeAnthropic
Claude Fable 5.1's 1M-token frontier lane, Sonnet 5's permanent launch pricing, cheaper cache reads, and Azure CCU billing on Microsoft Foundry.
from $2.00 /M in→
GeminiGoogle
Gemini 3.7 Flash is the new flagship, and Gemini 3.6 Flash's price is halved to match it — both $0.75/$0.075/$3.75 per M through year-end.
from $0.75 /M in→
GrokxAI
Grok 4.6 is direct-API only for now; Microsoft Foundry resells the 4.x line up to Grok-4.3, which now carries a published cache meter.
from $0.20 /M in→
QwenAlibaba
Qwen3.8 Max is the flagship, while Qwen3.8 Flash adds a new $0.15/$0.47 multimodal direct lane — but Foundry still has no per-token Qwen meter.
from $0.10 /M in→
MistralMistral AI
Mistral Medium 3.5 is now on Microsoft Foundry at the same rate as Mistral's own API.
from $1.50 /M in→
MiniMaxMiniMax AI
MiniMax M2.5 and MiniMax 3 are resold on Microsoft Foundry as Data Zone-only serverless listings; the direct API adds M2.7 and M3 as the current lineup for comparison.
from $0.30 /M in→
Embeddings
The retrieval layer. Input-only pricing, and the cheapest model is rarely the right one for code RAG.
from $0.02 /M in→
The framework
Don't trust the sticker price
Seven things that decide what you actually pay. None of them is the headline number.
Reusable method
How to verify real LLM costs
Six steps for finding what a workload really costs, including how to back out an undocumented cache rate from a billing export.
- 1
Get the provider's published per-million rate for input, output, and cached input if it's listed.
- 2
Run a real workload and read the actual billed cost from Azure Cost Management, not a chat client's estimate.
- 3
If real cost is far below the flat-rate math, suspect an undocumented cache discount before assuming a bug.
- 4
Export a cost report grouped by meter, not by service name. Match known meters to confirmed rates to back out token counts, then solve for the unknown:
cached_rate = cached_meter_cost / cached_token_count. - 5
Cross-check those token counts against the usage dashboards to validate your assumptions.
- 6
Treat anything not on an official page as an estimate, not fact. Re-check once real numbers land or a billing cycle closes.
Confidenceofficial published pagederived reconciled from billing †estimate pattern-inferred ‡