claude-fable-5 vs grok-build-0.1

Pricing, Performance & Features Comparison

claude-fable-5

Authoranthropic

Context Length1M

Reasoning

Providers1

ReleasedJun 2026

Knowledge Cutoff–

License–

Anthropic’s most capable widely released model, a tier above Opus. Thinking is always on, sampling parameters are rejected, and requests may return a refusal stop reason.

Input$10

Output$50

Latency (p50)8.1s

Output Limit128K

Function Calling

JSON Mode

InputText, Image

OutputText

anthropic

in$10out$50cache$1write$13

Latency (24h)

Success Rate (24h)

grok-build-0.1

Authorxai

Context Length256K

Reasoning

Providers1

ReleasedMay 2026

Knowledge Cutoff–

License–

xAI’s agentic coding model and the successor to grok-code-fast-1; also suited to inexpensive general tool-calling.

Input$1

Output$2

Latency (p50)5.7s

Output Limit–

Function Calling

JSON Mode

InputText, Image

OutputText

xai

in$1out$2cache$0.2–