anomalyco--models.dev
wehub-resource-sync
107 行
2.4 KiB
TOML
107 行
2.4 KiB
TOML
name = "Claude Sonnet 4.6"
|
|
description = "Claude workhorse for coding agents, careful analysis, and production cost control"
|
|
family = "claude-sonnet"
|
|
release_date = "2026-02-17"
|
|
last_updated = "2026-03-13"
|
|
attachment = true
|
|
reasoning = true
|
|
temperature = true
|
|
tool_call = true
|
|
knowledge = "2025-08-31"
|
|
open_weights = false
|
|
|
|
[limit]
|
|
context = 1_000_000
|
|
output = 64_000
|
|
|
|
[modalities]
|
|
input = ["text", "image", "pdf"]
|
|
output = ["text"]
|
|
|
|
[[benchmarks]]
|
|
name = "SWE-Atlas Codebase QnA"
|
|
score = 31.2
|
|
metric = "score"
|
|
harness = "Claude Code"
|
|
source = "https://labs.scale.com/leaderboard/sweatlas-qna"
|
|
|
|
[[benchmarks]]
|
|
name = "SWE-Atlas Refactoring"
|
|
score = 32.21
|
|
metric = "score"
|
|
harness = "Claude Code"
|
|
source = "https://labs.scale.com/leaderboard/sweatlas-refactoring"
|
|
|
|
[[benchmarks]]
|
|
name = "SWE-Atlas Test Writing"
|
|
score = 31.76
|
|
metric = "score"
|
|
harness = "Claude Code"
|
|
source = "https://labs.scale.com/leaderboard/sweatlas-tw"
|
|
|
|
[[benchmarks]]
|
|
name = "Artificial Analysis Coding Agent Index"
|
|
score = 49.4
|
|
metric = "average pass@1"
|
|
harness = "Claude Code"
|
|
variant = "medium"
|
|
source = "https://artificialanalysis.ai/agents/coding-agents"
|
|
|
|
[[benchmarks]]
|
|
name = "SWE-Atlas Codebase QnA"
|
|
score = 70.3
|
|
metric = "pass@1"
|
|
harness = "Claude Code"
|
|
variant = "medium"
|
|
source = "https://artificialanalysis.ai/agents/coding-agents"
|
|
|
|
[[benchmarks]]
|
|
name = "SWE-Bench Pro"
|
|
score = 14.9
|
|
metric = "pass@1"
|
|
harness = "Claude Code"
|
|
variant = "medium"
|
|
dataset = "hard-aa"
|
|
source = "https://artificialanalysis.ai/agents/coding-agents"
|
|
|
|
[[benchmarks]]
|
|
name = "Terminal-Bench"
|
|
score = 63.1
|
|
metric = "pass@1"
|
|
harness = "Claude Code"
|
|
variant = "medium"
|
|
version = "2.1"
|
|
source = "https://artificialanalysis.ai/agents/coding-agents"
|
|
|
|
[[benchmarks]]
|
|
name = "Terminal-Bench"
|
|
score = 67.0
|
|
metric = "success rate"
|
|
harness = "Terminus-2"
|
|
version = "2.1"
|
|
source = "https://www.anthropic.com/news/claude-sonnet-5"
|
|
date = "2026-06-30"
|
|
|
|
[[benchmarks]]
|
|
name = "Humanity's Last Exam"
|
|
score = 34.6
|
|
metric = "accuracy"
|
|
variant = "no tools"
|
|
source = "https://www.anthropic.com/news/claude-sonnet-5"
|
|
date = "2026-06-30"
|
|
|
|
[[benchmarks]]
|
|
name = "Humanity's Last Exam"
|
|
score = 46.8
|
|
metric = "accuracy"
|
|
variant = "with tools"
|
|
source = "https://www.anthropic.com/news/claude-sonnet-5"
|
|
date = "2026-06-30"
|
|
|
|
[[benchmarks]]
|
|
name = "OSWorld-Verified"
|
|
score = 78.5
|
|
metric = "success rate"
|
|
source = "https://www.anthropic.com/news/claude-sonnet-5"
|
|
date = "2026-06-30"
|