Elorian: Glimpse vs StepFun: Step 5 Preview
A side-by-side comparison of Elorian: Glimpse ($2.10 / 1M tokens) and StepFun: Step 5 Preview ($3.50 / 1M tokens) using current USD prices and capabilities from the api14 catalog. Both run through the same OpenAI-compatible API.
Specifications
| Elorian: Glimpse | StepFun: Step 5 Preview | |
|---|---|---|
| Provider | Elorian | StepFun |
| Input | $2.10 / 1M tokens | $3.50 / 1M tokens |
| Output | $10.50 / 1M tokens | $9.45 / 1M tokens |
| Cache read | $0.21 / 1M tokens | $0.35 / 1M tokens |
| Context window | 262,144 | 1,048,576 |
| Max output | 256,000 | 64,000 |
| Input | text, image, video | text, image, video |
| Output | text | text |
| Tool calling | ||
| Structured output | ||
| Reasoning control | ||
| Added | Oct 17, 2026 | Oct 8, 2026 |
Key differences
- Elorian: Glimpse is 40% cheaper on input tokens.
- StepFun: Step 5 Preview is 10% cheaper on output tokens.
- StepFun: Step 5 Preview reads 4.0× more context (1,048,576 tokens).
- A request with 1,000 input and 500 output tokens costs $0.00735 with Elorian: Glimpse and $0.00823 with StepFun: Step 5 Preview.
In depth
Elorian: Glimpse is a text model from Elorian, released on 2026-10-17. StepFun: Step 5 Preview is a text model from StepFun, released on 2026-10-08. Both are served through api14 under the model ids elorian/glimpse and stepfun/step-5-preview, and both accept text, image and video as input while producing text output only. Neither is a specialist in a single narrow task; they sit in the same general category, so the choice between them comes down to context size, parameter support and price rather than to a fundamental difference in what they do.
The clearest split is context length. Step 5 Preview offers a context window of 1,048,576 tokens, four times the 262,144 tokens of Glimpse, so it is the model to reach for when you need to keep very large amounts of material in a single call, such as long transcripts, whole repositories of documents or extended video input alongside text. On modalities the two are identical, since both take text, image and video and both return text. Their parameter lists overlap heavily: both support temperature, top_p, top_k, min_p, frequency and presence penalties, repetition penalty, stop, max_tokens, logprobs, top_logprobs, response_format, structured_outputs, tools, tool_choice, reasoning and include_reasoning. Step 5 Preview adds reasoning_effort and seed, while Glimpse adds logit_bias. If you depend on seeding for reproducible sampling, or on reasoning_effort to tune how much internal reasoning the model does, that points to Step 5 Preview; if you rely on logit_bias to steer token probabilities, that points to Glimpse.
In practice, pick Step 5 Preview when the job is dominated by volume of input: multi hour video review, large document sets, or agent loops that accumulate a long history and cannot afford to truncate. Pick Glimpse when your inputs fit comfortably within a quarter of a million tokens and you want the cheaper input side, since Glimpse is the cheaper of the two on input pricing. Output pricing runs the other way, with Step 5 Preview cheaper on output, so workloads that generate a lot of text relative to what they read may end up cheaper on Step 5 Preview even before context is considered. For mixed workloads, the two are close enough in price that the deciding factor is usually which side of the input and output balance your traffic falls on, plus whether you need the extra parameters.
Calling either model through api14 is the same operation. The API is OpenAI compatible, so you point your existing client at the api14 base URL, send your key, and set the model field to elorian/glimpse or stepfun/step-5-preview. Everything else in the request body, messages, temperature, tools, response_format and the rest, keeps the shape you already use. Switching between the two is a one line change to the model id, which makes it cheap to route a request to Glimpse for smaller inputs and to Step 5 Preview for the very large ones, or to A/B the two on the same prompt without rewriting your integration.
Switch with one string
Both models use the same endpoint and request shape. Change the model value from "elorian/glimpse" to "stepfun/step-5-preview" and prompts, streaming and tools keep working.
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": "stepfun/step-5-preview",
"messages": [
{
"role": "user",
"content": "Explain vector databases in two sentences."
}
]
});
console.log(result.choices[0].message.content);
More comparisons
- LiquidAI: LFM3-1.7B vs Ricursive: Ripple 65B Preview
- LiquidAI: LFM3-1.7B vs Crouton Alpha
- LiquidAI: LFM3-1.7B vs Tencent: Hy4
- LiquidAI: LFM3-1.7B vs Z.ai: GLM 5.3 FlashX Legal
- Ricursive: Ripple 65B Preview vs Crouton Alpha
- Ricursive: Ripple 65B Preview vs Tencent: Hy4
- Ricursive: Ripple 65B Preview vs Z.ai: GLM 5.3 FlashX Legal
- Crouton Alpha vs Tencent: Hy4