Claude Sonnet 5
Model at Anthropic
The basics
As of Nov 1- Status
- Price
- Context
- Open weights
- Speed
Updates
Role
Mid-size model at Anthropic
Sep 2026 to now · 2 months
Jun 2026 to Sep 2026 · 3 months
Feb 2026 to Jun 2026 · 4 months
Works with
From recorded changes
Agent · 4 shared changes
App · 3 shared changes
API · 3 shared changes
Agent · 3 shared changes
Similar models
$2 in, $10 out
$3 in, $15 out
$3 in, $15 out
Cost vs peers
Tied for cheapest of 4 similar models by output price.
Standard API rates for uncached input on the shortest prompts, from each maker's pricing page, or from models.dev for models not researched here yet.
Where it works
Prices and uptime are OpenRouter's figures for each provider's cheapest endpoint. Uptime covers the last 24 hours.
Benchmarks
Higher is better. Green is Claude Sonnet 5. Dark marks are similar models.
Independent
What people say
Reviews (1)
Lior GrossmanShipping software since 1993, now building with AISonnet 5 wrote hundreds of nonfiction books for one of my products. In my August 2026 blind reads against Sonnet 4.6, most chapters were pretty good on both sides, and Sonnet 5 was mostly good and slightly better: more interesting, more concrete, and easier to understand, with fewer formulas. I kept it over Opus 5, which scored higher with an AI judge at about 2.5 times the cost per book. The effort setting matters for writing prose. The default, high, was the worst on quality, cost and speed, and sometimes it thought so long that it stopped a chapter mid-sentence. Low wrote books that were just as good, for about 40% less. For business decisions I still want Opus.
Oct 31, 2026
Launch quotes, picked by Anthropic
From its launch post, so not a neutral sample.
"Claude Sonnet 5 gives our agents a strong execution layer for multi-step software engineering work. It handles sustained coding, tool use, and debugging well across messy technical contexts, and has been especially useful for workflows where follow-through and technical grounding matter."
"We handed Claude Sonnet 5 a two-part job—update Salesforce account tiers, send a launch announcement to enterprise contacts—and it finished end to end. That used to stall halfway. For day-to-day automation, it’s a no-brainer."
Incidents
Incidents per week, last 13 weeks. Most in one week: 2.
- Sep 3Anthropic reported elevated errors for Claude Sonnet 519 min
- Sep 2Anthropic reported elevated errors for Claude Sonnet 514 min
- Aug 18Anthropic reported elevated errors on several Claude models2 h 41 min
- Aug 13Anthropic reported elevated errors on three Claude models1 h 35 min
- Aug 5Anthropic reported elevated errors on four Claude models7 h 9 min
Writing habits
Creative writing · EQ-Bench, among 51 models tracked here
Slop score
11.6
Overused AI words and phrases. Less slop than 72% of tracked models.
Repetition
4.6
Repeated words and ideas. Less repetition than 16% of tracked models.
Story length
5,753
Average characters per story. Longer than 10% of tracked models.