Check whether premium spend matches expected users and agents
The page should make surprising model selection stand out quickly.
QI
Models Detail
A model detail page should pair provider facts with the human and session patterns that make the lane operationally important.
Use This Page When Asking
Why does this model lane matter beyond cost, and how does it change external-provider exposure?
Open c6 raw detailSessions
5
Small but high-consequence lane
Spend
$38.10
Top model total
Cache hit rate
62%
Meaningful cost signal
Top user
bob@company.com
Lane anchored to one operator
Why This Detail Matters
Important Metadata
Model ID
claude-sonnet-4-6
Family
Sonnet
Provider
Anthropic
Provider Host
api.anthropic.com
Base URL
https://api.anthropic.com/v1/messages
Context Window
200k tokens
What This Detail Should Help Decide
Compare cache behavior against cost before escalating
A high-spend lane can still be routine if the work shape and cache profile fit expectation.
Keep provider context attached to the model story
The reviewer should not have to reconstruct the external boundary from memory.
Cost And Usage Shape
Spend Concentration
A small session count drives a disproportionate share of value and cost
This model family is not the broadest lane, but it is one of the most consequential because a few premium sessions account for an outsized share of spend.
Token Economics
Pricing belongs close to the model story
Input, output, and cache-read pricing are part of what makes this lane operationally distinct. Reviewers should not have to jump away from the model page to understand that economic profile.
Cache Read
Cache behavior changes the meaning of high spend
A sixty-two percent cache hit rate softens the raw cost story. Premium spend is more alarming when cache efficiency is weak or drifting, and more ordinary when the work shape still fits expectation.
Model-Centered Topology
A model map should pair usage footprint with spend, cache efficiency, and the workflows it powers.
Useful per model: provider/family, who uses it, which agents and keys select it, what session shapes it appears in, and whether spend or cache behavior is changing.
Model
claude-opus-4-6
Premium model concentrated in a small set of high-cost sessions.
5 sessions · $38.10 total
Provider
Anthropic
Family
claude-opus
Cache hit rate
62%
Spend posture
top spend model
Primary Consumers
Who selects this model most often.
bob@company.comtop user
Cursor v0.42.0top agent
sk-ant-...A3F2credential
Session Shape
The type of work this model is serving.
long coding sessions44 turns max
high token outputpremium spend
cache reads62%
Comparative Context
How it differs from adjacent models.
claude-sonnet-4-6volume leader
claude-haiku-4-5CI / cheap path
gpt-4orare fallback
Watch For
The questions worth asking before cost spikes.
cost slope$38.10 total
new provider mixrouting shift
cache dropefficiency loss
Representative Event Chain
The page should show how the item participates in a readable sequence, not just a pile of supporting rows.
- 1
Selection
A small group of users repeatedly choose the model for a specific work shape
The lane first becomes legible through the people and agents that select it most often.
bob@company.comCursorclaude-opus-4-6 - 2
Work pattern
Long turns and higher output explain the cost profile
The page should relate spend to the kind of work the model is actually powering.
44 turnshigh outputpremium spend - 3
Boundary meaning
Provider traffic turns model choice into an external context question
The model lane matters partly because it is a repeated outbound path for sensitive company context.
Anthropicapi.anthropic.com - 4
Confirmation
The reviewer pivots into pricing and recent sessions after the lane is framed
Once the model lane is understood, pricing rows and raw sessions become confirmation tools.
pricingrecent sessions
c6 Evidence Tables
Pricing
c6 keeps model economics close to model identity.
| Token type | Price per MTok (USD) |
|---|---|
| input | $3.00 |
| output | $15.00 |
| cache read | $0.30 |
Recent Sessions
The raw page ties the model back to users and agents.
| User | Agent | Turns | Cost | Status |
|---|---|---|---|---|
| alice@company.com | Claude Code | 31 | $8.20 | ended |
| carol@company.com | Claude Code | 5 | $0.28 | active |
| ci-runner-04 | LangChain | 2 | $0.02 | active |
Return To Category
Compare this representative detail against the full category framing and drill targets.
Open Models overviewOpen c6 Raw Detail
Use the original detail page for exact raw evidence after the c12 story framing.
Open c6 raw detailOpen c6 Raw List
Compare this item against the broader c6 entity ledger once the representative story is clear.
Open c6 raw list