Model Comparator

Compare up to four models using the same criteria.

Qwen3.8 27B (Medium)
Alibaba
Grok Build 0.1 0616
SpaceXSI

Differences only

✓ Best published value among measured models. — Data unavailable. These metrics do not establish an overall ranking.

Model Comparator
Metrics	Qwen3.8 27B (Medium) Alibaba	Grok Build 0.1 0616 SpaceXSI
Intelligence	Best: 27.6	27.2
Coding	Best: 56.1	51.5
Agentic	44.4	—
Speed	43.2 t/s	Best: 67.4 t/s
TTFT	3.96 s	Best: 1.76 s
Input · $ / 1M tokens	Best: $0.50	$1×2 relative to the lowest price among the compared models
Output · $ / 1M tokens	$3×1.5 relative to the lowest price among the compared models	Best: $2
Context window	256K	256K
Reasoning	Yes	Yes
Open weights	Yes	No
Cost / task	$1.13	—

Intelligence

Qwen3.8 27B (Medium): Best: 27.6
Grok Build 0.1 0616: 27.2

Coding

Qwen3.8 27B (Medium): Best: 56.1
Grok Build 0.1 0616: 51.5

Agentic

Qwen3.8 27B (Medium): 44.4
Grok Build 0.1 0616: —

Speed

Qwen3.8 27B (Medium): 43.2 t/s
Grok Build 0.1 0616: Best: 67.4 t/s

TTFT

Qwen3.8 27B (Medium): 3.96 s
Grok Build 0.1 0616: Best: 1.76 s

Input · $ / 1M tokens

Qwen3.8 27B (Medium): Best: $0.50
Grok Build 0.1 0616: $1×2 relative to the lowest price among the compared models

Output · $ / 1M tokens

Qwen3.8 27B (Medium): $3×1.5 relative to the lowest price among the compared models
Grok Build 0.1 0616: Best: $2

Context window

Qwen3.8 27B (Medium): 256K
Grok Build 0.1 0616: 256K

Reasoning

Qwen3.8 27B (Medium): Yes
Grok Build 0.1 0616: Yes

Open weights

Qwen3.8 27B (Medium): Yes
Grok Build 0.1 0616: No

Cost / task

Qwen3.8 27B (Medium): $1.13
Grok Build 0.1 0616: —

Each reasoning level remains a separate configuration.

Score and cost

Each line joins the reasoning levels of a compared model; the large point marks the configuration in the table. The top-right corner holds the best score for the lowest cost. Select another point to compare that level.

Qwen3.8 27B (Medium)27.6 · $1.13
Grok Build 0.1 061627.2 · $1.25