Model Comparator

Compare up to four models using the same criteria.

Claude 4.5 Haiku (Non-reasoning)
Anthropic
Step 3.5 Flash
StepFun

Differences only

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

Model Comparator
Metrics	Claude 4.5 Haiku (Non-reasoning) Anthropic	Step 3.5 Flash StepFun
Intelligence	16.9	Best: 17
Coding	43.9	—
Math	39	—
Agentic	8	—
Speed	81.9 t/s	Best: 122 t/s
TTFT	Best: 0.58 s	3.28 s
Input · $ / 1M tokens	$1×10 relative to the lowest price among the compared models	Best: $0.10
Output · $ / 1M tokens	$5×16.7 relative to the lowest price among the compared models	Best: $0.30
Context window	1M	1M
Reasoning	No	Yes
Open weights	No	Yes

Intelligence

Claude 4.5 Haiku (Non-reasoning): 16.9
Step 3.5 Flash: Best: 17

Coding

Claude 4.5 Haiku (Non-reasoning): 43.9
Step 3.5 Flash: —

Math

Claude 4.5 Haiku (Non-reasoning): 39
Step 3.5 Flash: —

Agentic

Claude 4.5 Haiku (Non-reasoning): 8
Step 3.5 Flash: —

Speed

Claude 4.5 Haiku (Non-reasoning): 81.9 t/s
Step 3.5 Flash: Best: 122 t/s

TTFT

Claude 4.5 Haiku (Non-reasoning): Best: 0.58 s
Step 3.5 Flash: 3.28 s

Input · $ / 1M tokens

Claude 4.5 Haiku (Non-reasoning): $1×10 relative to the lowest price among the compared models
Step 3.5 Flash: Best: $0.10

Output · $ / 1M tokens

Claude 4.5 Haiku (Non-reasoning): $5×16.7 relative to the lowest price among the compared models
Step 3.5 Flash: Best: $0.30

Context window

Claude 4.5 Haiku (Non-reasoning): 1M
Step 3.5 Flash: 1M

Reasoning

Claude 4.5 Haiku (Non-reasoning): No
Step 3.5 Flash: Yes

Open weights

Claude 4.5 Haiku (Non-reasoning): No
Step 3.5 Flash: Yes

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.

Claude 4.5 Haiku (Non-reasoning)16.9 · $2
Step 3.5 Flash17 · $0.15