modelbenchmark.io

Claude Sonnet 5.5

Anthropic · claude-sonnet-5-5

Compare

Fast Claude model for everyday coding, agents, and knowledge work

Specification

anthropic listing

Context
1M
Max output
128K
Released
2026-09-28
Knowledge cutoff
2026-06
Retires
—
Open weights
no
Input
text, image, pdf
Output
text

Price

US dollars per million tokens · anthropic list price

Input
$2.00
Output
$10.00
Cache read
$0.20
Cache write
$2.50

Batch and priority prices

US dollars per million tokens · per listing

PriceListing$/M
Batch inputanthropic$1.00
Batch outputanthropic$5.00
Cache write, 1 houranthropic$4.00
Cache write, 1 hourbedrock_mantle · bedrock-mantle$4.40
Cache write, 1 hourbedrock_mantle · bedrock-mantle$4.80
Cache write, 1 houropenrouter · batch$2.00

Batch input is 50% below the interactive input price at the same listing.

Quality

8 benchmarks · 2 sources

Composite
97th
percentile of 340 scored models
Rank
13 / 340
±0.08 sd
Evidence
8 × 2
3 or more benchmarks, from 2 or more sources
Effort range
2.4
points between effort settings
reasoning98th2/4 bench
math96th3/7 bench
knowledge59th1/1 bench
general52nd1/1 bench

By reasoning effort

every setting placed on the same scale as the leaderboard

SettingCompositeBenchmarks behind it
max97th8
xhigh81st1

Every score

one row per source and configuration — nothing averaged away

BenchmarkScoreConfigurationSourceRun
Mystery Game Puzzles65.0 ±4.8maxEpoch AI2026-09-29
Furniture Assembly75.0 ±5.5maxEpoch AI2026-09-29
FrontierMath-Tiers-1-3-v288.8 ±1.9maxEpoch AI2026-09-29
FrontierMath-Tier-4-v280.5 ±6.3maxEpoch AI2026-09-29
GPQA diamond95.6 ±1.4maxEpoch AI2026-09-29
OTIS Mock AIME 2024-2025100.0 ±0.0maxEpoch AI2026-09-29
LiveBench78.2xhigh · 23 runsLiveBench2026-06-25
↳75.8max · 23 runsLiveBench2026-06-25
Settings differ by 2.4 points on this benchmark.
SimpleQA Verified46.5 ±1.6maxEpoch AI2026-09-29

The leading figure for a benchmark is the median across its configurations, so one heroic high-effort run cannot set the number.

Available from 32 hosts

HostIn $/MOut $/MCache rdCache wrContextOutputRetires
anthropicvendor$2.00$10.00$0.20$2.501M128K—
openrouter · batch$1.00$5.00$0.10$1.251M128K—
amazon-bedrock · global$2.00$10.00$0.20$2.501M128K—
azure$2.00$10.00$0.20$2.501M128K—
azure_ai$2.00$10.00$0.20$2.501M128K—
azure-cognitive-services$2.00$10.00$0.20$2.501M128K—
bedrock_converse$2.00$10.00$0.20$2.501M128K—
bedrock_converse · global$2.00$10.00$0.20$2.501M128K—
crossmodel$2.00$10.00$0.20$2.501M128K—
edenai$2.00$10.00$0.20$2.501M128K—
github-copilot$2.00$10.00$0.20$2.501M128K—
kilo$2.00$10.00$0.20$2.501M128K—
llmgateway$2.00$10.00$0.20$2.501M128K—
llmgateway-providers$2.00$10.00$0.20$2.501M128K—
merge-gateway$2.00$10.00$0.20$2.501M128K—
nano-gpt$2.00$10.00$0.20$2.501M128K—
nano-gpt · thinking$2.00$10.00$0.20$2.501M128K—
ofox$2.00$10.00$0.20$2.501M128K—
opencode$2.00$10.00$0.20$2.501M128K—
openrouter$2.00$10.00$0.20$2.501M128K—
perplexity$2.00$10.00$0.20————
requesty$2.00$10.00$0.20$2.501M128K—
vercel$2.00$10.00$0.20$2.501M128K—
vertex_ai-anthropic_models$2.00$10.00$0.20$2.501M128K—
bedrock_converse · apac$2.20$11.00$0.22$2.751M128K—
bedrock_converse · au$2.20$11.00$0.22$2.751M128K—
bedrock_converse · eu$2.20$11.00$0.22$2.751M128K—
bedrock_converse · jp$2.20$11.00$0.22$2.751M128K—
bedrock_converse · us$2.20$11.00$0.22$2.751M128K—
bedrock_mantle · bedrock-mantle
$2.20
$2.40
$11.00
$12.00
$0.22
$0.24
$2.75
$3.00
1M128K—
bedrock$2.40$12.00$0.24$3.001M128K—
venice$3.75$18.75$0.375$4.68751M128K—

More from Anthropic

most-hosted first