MDL
Claude 4.5 Sonnet
Anthropic
Verified against Artificial Analysis · Jul 13, 2026
Capability
84
percentile index
Price /M
$3 / $15
in / out
Context
—
Avg score
62.9
10 benchmarks
AA Intelligence
36.4
index
Speed
52 tok/s
TTFT 9.6s
Released
Sep 29, 2025
About
Claude 4.5 Sonnet is a proprietary model in the Claude family from Anthropic. Benchmarked on 10 evals, averaging 62.9.
Benchmark scores · 10
avg 62.9 — higher is betterThe research behind this model
all papers mentioning it →OmniOPD: Logit-Free On-Policy Distillation via Speculative Verification▲ 8 on HF · May 31, 2026Enhancing Reliability in LLM-Based Secure Code GenerationMay 22, 2026Evaluating Commercial AI Chatbots as News IntermediariesMay 21, 2026NodeSynth: Socially Aligned Synthetic Data for AI EvaluationMay 14, 2026Magis-Bench: Evaluating LLMs on Magistrate-Level Legal TasksMay 8, 2026
Mentions matched by name in title/abstract — from the arXiv + HF daily corpus.
Performance & price
Output speed52 tok/s
Latency (TTFT)9.6s
Price in / out$3 / $15 per 1M
Blended price$6 / 1M
Coding index52.1
Math index88.0
Median across API providers, via Artificial Analysis. Blended = 3:1 input:output. For reasoning models, latency includes thinking time. Methodology →
History
full ledger →Capability over time
Append-only ledger — every observed change to this model's numbers.
API providers
Artificial Analysis$15/M