Model

Grok-1.5

CLOSED

SpaceXAI · Grok 1.5 family · released Mar 29, 2024

First major closed xAI iteration with 128K context; never exposed via a public developer API (xAI's API launched later, atop Grok-2).

ReasoningCodingVisionFunction callingTool useAgentic

808.7

Elo · rank #252

Parameters

Undisclosed

Active params

Undisclosed

Context

128K tokens

Architecture

Mixture-of-Experts (exact configuration undisclosed by xAI)

License

Proprietary

Languages

—

API price (in/out)

No hosted API

Modalities

text

Benchmark results

Bar shows position within the tracked field; marker = field best

GPQA DiamondReasoning35.9%#238

best: GPT-5.6 · 94.6%

GSM8KMath90.0%#41

best: Llama 3.1 405B · 96.8%

HumanEvalCoding74.1%#69

best: Claude Opus 4.5 · 99.4%

MATH-500Math50.6%#130

best: GPT-5 · 99.4%

MMLU-ProKnowledge51.0%#132

best: Claude Fable 5 · 91.5%

MMLUKnowledge81.3%#78

best: OpenAI o3 · 92.9%

Run it locally

Closed weights — available via API only. No local deployment.

Input / M tok

—

Output / M tok

—

Grok 1.5 family

Elo progression across releases

Grok-1.5Mar 2024808.7

Resources

—

API price — · each benchmark row carries its own source badge (see methodology)