ModelsCompareBest forBenchmarksStatusPricingAPI

Claude Opus 4.8

Anthropic·Operational·Released 2026-05-28textvision
Context1,000k
Max output128k
Input / 1M$5 ↗
Output / 1M$25 ↗
Cached in$0.5
Benchmarks17

Benchmarks

Cited public sources · rank is exact-benchmark, apples-to-apples

AIME 2026 (MathArena)#1/21100% ↗
FrontierMath Tier 4 v2 (Epoch AI run)#10/4456.1% ↗
FrontierMath Tiers 1-3 v2 (Epoch AI run)#10/5980% ↗
GPQA Diamond (Epoch AI run)#19/7791% ↗
OTIS Mock AIME 2024-2025 (Epoch AI run)#9/7698.3% ↗
SimpleQA Verified (Epoch AI run)#12/5353% ↗
LMArena Elo#14/621452.7 ↗
BrowseComp#4/1584.3% ↗
GPQA Diamond#7/7093.6% ↗
Humanity's Last Exam (no tools)#3/2949.8% ↗
Humanity's Last Exam (with tools)#3/1357.9% ↗
OSWorld-Verified#2/1283.4% ↗
SWE-bench Multilingual#1/784.4% ↗
SWE-bench Multimodal#1/238.4% ↗
SWE-Bench Pro#1/1569.2% ↗
SWE-bench Verified#1/4588.6% ↗
Terminal-Bench 2.1#8/1774.6% ↗

Compare

vs GPT-6 Astravs GPT-5.6 Solvs GPT-5.6 Terravs GPT-5.6 Lunavs Claude Fable 5.1vs Claude Opus 5vs Claude Sonnet 5vs Claude Haiku 4.5

More from Anthropic

Claude Fable 5.1Claude Opus 5Claude Sonnet 5Claude Haiku 4.5Claude Opus 4.7Claude Opus 4.6

Access Claude Opus 4.8 and every other model through one endpoint with automatic failover: Respan gateway.