Compare models

Context, limits, capabilities, price and use cases side by side — up to 4 models at a time. The URL holds your selection, so you can share it.

Claude Opus 4.8Claude Opus 5 Add a modelChat with these 2

AttributeHide identical rows	Claude Opus 4.8anthropic/claude-opus-4-8	Claude Opus 5anthropic/claude-opus-5
Specs
Context window	1M	1M
Max output	128K	128K
Input modalities	TextImage	TextImage
Output modalities	Text	Text
Capabilities
Streaming	Supported	Supported
Tool calling	Supported	Supported
Structured output (JSON)	Supported	Supported
Image input	Supported	Supported
Thinking mode	Supported	Supported
Price / 1M tokens
Input	$ 5.00	$ 5.00
Output	$ 25.00	$ 25.00
Cache read	$ 0.50	$ 0.50
Cache write	$ 6.25	$ 6.25
What it is good at
Best for	Agentic coding Security research with reduced guardrails	Agentic coding Professional knowledge work Scientific research
Published benchmarks
Frontier-Bench v0.1score	—	43.3%(setup not fully disclosed)
FrontierCode v1.1 (Main)score	—	53.4%(setup not fully disclosed)
Terminal-Bench 2.1pass_rate	74.6%(setup not fully disclosed)	—
BrowseComp (with tools)accuracy	—	90.8%(setup not fully disclosed)
AutomationBenchscore	—	26%(setup not fully disclosed)
SWE-Bench Proresolved_rate	69.2%(setup not fully disclosed)	—
OSWorld 2.0success_rate	—	70.6%(setup not fully disclosed)
OSWorld-Verifiedsuccess_rate	83.4%(setup not fully disclosed)	—
Finance Agent v2score	53.9%(setup not fully disclosed)	—
GDPval-AAelo	1890(setup not fully disclosed)	—
GDPval-AA v2elo	—	1861(setup not fully disclosed)
ARC-AGI-3score	—	30.2%(setup not fully disclosed)
Humanity's Last Examaccuracy	49.8%(setup not fully disclosed)	56.3%(setup not fully disclosed)
Humanity's Last Exam (with tools)accuracy	57.9%(setup not fully disclosed)	64.7%(setup not fully disclosed)