We use cookies. Cookies and browser storage keep you signed in and remember your preferences. With your permission we also measure how the site is used. You can any time. See our cookie policy.

New here? See how it works A 1-minute tour of the playground, models and your API key.

StepFun: Step 5 Preview vs SpaceXAI: Grok 4.3

A side-by-side comparison of StepFun: Step 5 Preview ($3.50 / 1M tokens) and SpaceXAI: Grok 4.3 ($4.38 / 1M tokens) using current USD prices and capabilities from the api14 catalog. Both run through the same OpenAI-compatible API.

Specifications

	StepFun: Step 5 Preview	SpaceXAI: Grok 4.3
Provider	StepFun	xAI
Input	$3.50 / 1M tokens	$4.38 / 1M tokens
Output	$9.45 / 1M tokens	$8.75 / 1M tokens
Cache read	$0.35 / 1M tokens	$0.70 / 1M tokens
Web search	-	$17.50 / 1K searches
Context window	1,048,576	1,000,000
Max output	64,000	900,000
Input	text, image, video	text, image, file
Output	text	text
Tool calling
Structured output
Reasoning control
Added	Oct 8, 2026	Apr 30, 2026

Key differences

StepFun: Step 5 Preview is 20% cheaper on input tokens.
SpaceXAI: Grok 4.3 is 7% cheaper on output tokens.
A request with 1,000 input and 500 output tokens costs $0.00823 with StepFun: Step 5 Preview and $0.00875 with SpaceXAI: Grok 4.3.

In depth

StepFun: Step 5 Preview is a text model from StepFun, released as a preview on 8 October 2026. SpaceXAI: Grok 4.3 is a text model from xAI, released on 30 April 2026. Both are served on api14 under the text category, and both take non text input and return text only, so neither is a speech or image generation model. The naming is slightly confusing: the provider field for Grok 4.3 says xAI, while the display name carries the SpaceXAI label, so if you are matching model ids to vendors in your own config, use the model id rather than the display name.

The two models are close in context length. Step 5 Preview offers 1,048,576 tokens, Grok 4.3 offers 1,000,000 tokens, and the qualitative comparison treats the difference as not meaningful, so you should not pick between them on context alone. The input modalities do differ. Step 5 Preview accepts text, image and video input, while Grok 4.3 accepts text, image and file input. If your pipeline needs to hand a video to the model, Step 5 Preview is the one that lists video support; if you need to pass a document as a file, Grok 4.3 lists file support. On parameters, Step 5 Preview exposes a wider set, including frequency_penalty, min_p, presence_penalty, repetition_penalty, top_k and logprobs alongside the reasoning and structured output controls. Grok 4.3 exposes a narrower set: reasoning, reasoning_effort, include_reasoning, logprobs, top_logprobs, structured_outputs, tools, tool_choice, temperature, top_p, max_tokens, seed and response_format. Both support tools and structured outputs, so neither has an exclusive claim there.

For task selection, start with the modality you actually need. Video understanding, or workloads where you want fine control over sampling through penalties and top_k, point to Step 5 Preview. Document or file based analysis, where the input arrives as a file rather than an image or video, points to Grok 4.3. On price, Step 5 Preview is the cheaper option for input and Grok 4.3 is the cheaper option for output, so the cost balance depends on your input to output ratio; a long context, short answer workload leans toward Step 5 Preview, while a short prompt with a long generated answer leans toward Grok 4.3. The live price table on this page has the current numbers, so check it against your own token mix rather than assuming one model wins overall.

Calling either model through api14 uses the same OpenAI compatible surface. You point your client at the api14 base URL, send a chat completions request with your messages, and set the model field. The only change between the two is that string: use stepfun/step-5-preview for Step 5 Preview and x-ai/grok-4.3 for Grok 4.3. Everything else in the request body, including tools, response_format and the reasoning parameters, stays where it is, though you should drop parameters the target model does not list, such as frequency_penalty or top_k when you switch to Grok 4.3, since those are not in its supported set. Because both models are text output only, your response parsing code does not need to branch.

Switch with one string

Both models use the same endpoint and request shape. Change the model value from "stepfun/step-5-preview" to "x-ai/grok-4.3" and prompts, streaming and tools keep working.

StepFun: Step 5 Preview details & playground →SpaceXAI: Grok 4.3 details & playground →

import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://api14.com/api/v1",
  apiKey: process.env.API14_API_KEY,
});

const result = await client.chat.completions.create({
  "model": "x-ai/grok-4.3",
  "messages": [
    {
      "role": "user",
      "content": "Explain vector databases in two sentences."
    }
  ]
});
console.log(result.choices[0].message.content);

More comparisons

Step 5 Preview vs Grok 4.3: API price, context & features · api14