Step 5 Preview
Step 5 Preview is StepFun's flagship model for agentic work, built on a sparse Mixture-of-Experts architecture (27B active / 600B total parameters). It performs strongly in software engineering and professional knowledge work, with particular strength in finance.
It is designed for extended tasks that span large codebases and documents, using tools and refining results over multiple steps.
- Organization
- StepFun (China)
- Family
- —
- Providers
- 21
- Context
- 1,000,000
- Output limit
- 1,000,000
- Knowledge
- —
- Release
- 2026-09-16
- Updated
- 2026-09-20
- Weights
- Open
- Input
- text, image/jpeg, image/png, image/gif, image/webp, video/mp4, video/quicktime, video/x-matroska
- Output
- text
- Capabilities
- —
Providers
Compare every cataloged offer for this model, including its provider, route, token limits, price, and capabilities.
| Provider | Model ID | Context | Output | Input / output · 1M | Reasoning | Tools | Structured | Details |
|---|---|---|---|---|---|---|---|---|
| AIHubMix | step-5-preview | 1,000,000 | 64,000 | — / — | — | — | — | View |
| AI/ML API | stepfun/step-5-preview | 1,000,000 | 64,000 | — / — | — | — | — | View |
| CommandCode | stepfun/Step-5-Preview | 1,000,000 | 1,000,000 | — / — | — | — | — | View |
| Deep Infra | stepfun-ai/Step-5-Preview | 1,048,576 | 1,000,000 | — / — | — | — | — | View |
| DevPass (LLM Gateway) | step-5-preview | 1,000,000 | 1,000,000 | — / — | — | — | — | View |
| Eden AI | deepinfra/stepfun-ai/Step-5-Preview | 1,048,576 | 1,000,000 | — / — | — | — | — | View |
| Eden AI | tencent/step-5-preview | 1,000,000 | 1,000,000 | — / — | — | — | — | View |
| EmpirioLabs AI | step-5-preview | 1,024,000 | 131,072 | — / — | — | — | — | View |
| Kilo Gateway | stepfun/step-5-preview | 1,000,000 | 64,000 | — / — | — | — | — | View |
| LLM Gateway | tencent/step-5-preview | 1,000,000 | 65,536 | — / — | — | — | — | View |
| NanoGPT | stepfun/step-5-preview | 1,048,576 | 65,536 | — / — | — | — | — | View |
| Novita AI | stepfun/step-5-preview | 1,000,000 | 64,000 | — / — | — | — | — | View |
| Nvidia | stepfun-ai/step-5-preview | 1,048,576 | 1,048,576 | — / — | — | — | — | View |
| OpenCode | step-5-preview-free | 1,000,000 | 65,536 | — / — | — | — | — | View |
| OpenCode Go | step-5-preview-free | 1,000,000 | 65,536 | — / — | — | — | — | View |
| OpenCode Go | step-5-preview | 1,000,000 | 65,536 | — / — | — | — | — | View |
| StepFunvia OpenRouter | stepfun/step-5-preview fp8 | 1,000,000 | 64,000 | $1.00 / $2.70 | Yes | Yes | Yes | View |
| Requesty | step-5-preview | 1,048,576 | 65,536 | — / — | — | — | — | View |
| StepFun (China) | step-5-preview | 1,000,000 | 65,536 | — / — | — | — | — | View |
| StepFun (Global) | step-5-preview | 1,024,000 | 65,536 | — / — | — | — | — | View |
| StepFun Step Plan (China) | step-5-preview | 1,000,000 | 65,536 | — / — | — | — | — | View |
| StepFun Step Plan (Global) | step-5-preview | 1,000,000 | 65,536 | — / — | — | — | — | View |
| Vercel AI Gateway | stepfun/step-5-preview | 1,000,000 | 1,000,000 | — / — | — | — | — | View |
Prices and limits apply to each offer. A dash means the value is not provided. OpenRouter offers show the hosting provider and routing channel separately.
Pricing
Compare token rates and cache charges across available routes and hosting configurations.
Lowest listed input and output prices across the offers above; they may belong to different providers. These are base rates, before conditional pricing, discounts or additional fees.
stepfun/step-5-previewObserved - Input
- $1.00
- Output
- $2.70
- Cache read
- $0.05
Hosting prices through OpenRouter All endpoints and cache rates
| Endpoint | Input / 1M | Output / 1M | Cache read / 1M | Cache write / 1M |
|---|---|---|---|---|
| StepFunstepfun/fp8 | $1.00 | $2.70 | $0.05 | — |
Each row is a hosting configuration. Open an offer in Providers for its full pricing conditions and additional fees. Endpoint observations may differ from the routing catalog's base price.
Performance
Review reported latency and output-throughput percentiles from the latest observed rolling window.
Time to first token · milliseconds
All latency percentiles 1 endpoints
| Endpoint | P50 | P75 | P90 | P99 |
|---|---|---|---|---|
| StepFunstepfun/fp8 | 3,449.5 ms | 7,436.5 ms | 21,431.9 ms | 92,469.48 ms |
Output throughput · tokens per second
All throughput percentiles 1 endpoints
| Endpoint | P50 | P75 | P90 | P99 |
|---|---|---|---|---|
| StepFunstepfun/fp8 | 42 t/s | 93 t/s | 114 t/s | 151.09 t/s |
Percentiles describe the observed request distribution. These are snapshots of a rolling window, not a historical time series.
Uptime
Compare successful request rates across the latest 5-minute, 30-minute, and 24-hour observations.
Reported by OpenRouter at the time of collection; rate-limited requests are excluded. A dash means no measurement was reported. This is not a live availability check.
Benchmarks
Review attributed evaluation results on each source's original scale.
Sources: Artificial Analysis and Design Arena, via OpenRouter. Scores retain their original scales and are not OpenAI Suite ratings.
Detailed evaluations
| Evaluation | Configuration | Source | Results | Details |
|---|---|---|---|---|
| Composite indices | Step 5 Previewstepfun/step-5-preview-20261008 | Artificial Analysis | Intelligence 28.5 · Coding — · Agentic — | Details |
Sources: Artificial Analysis, Design Arena and OpenRouter. Evaluation configurations and scales differ; results are not interchangeable. Updated Nov 14, 2026.
Activity
Daily prompt and completion tokens reported for this model's routes in OpenRouter's top 50. Missing days have no published value.
632,191,853,921 tokens
2,587,364,754,135 tokens
3,149,891,365,259 tokens
2,466,863,011,089 tokens
3,249,049,356,669 tokens
2,966,448,461,113 tokens
3,737,466,999,284 tokens
1,325,113,004,745 tokens
179,047,849,184 tokens
566,291,253,816 tokens
234,728,339,316 tokens
640,127,467,179 tokens
584,527,912,314 tokens
799,960,228,162 tokens
800,503,370,026 tokens
800,963,159,992 tokens
801,693,051,154 tokens
960,227,433,475 tokens
1,177,246,450,717 tokens
959,885,814,181 tokens
799,697,698,812 tokens
747,826,323,105 tokens
1,011,767,065,339 tokens
1,019,617,389,126 tokens
1,018,531,639,718 tokens
1,108,844,042,845 tokens
1,121,266,385,272 tokens
961,674,015,765 tokens
959,480,801,893 tokens
963,487,289,365 tokens
803,264,228,801 tokens
961,165,982,850 tokens
963,033,631,845 tokens
959,517,302,632 tokens
Daily observations · 34 days
| Date (UTC) | Reported tokens | Reported routes |
|---|---|---|
| Nov 13, 2026 | 959,517,302,632 | 1 |
| Nov 12, 2026 | 963,033,631,845 | 1 |
| Nov 11, 2026 | 961,165,982,850 | 1 |
| Nov 10, 2026 | 803,264,228,801 | 1 |
| Nov 9, 2026 | 963,487,289,365 | 1 |
| Nov 8, 2026 | 959,480,801,893 | 1 |
| Nov 7, 2026 | 961,674,015,765 | 1 |
| Nov 6, 2026 | 1,121,266,385,272 | 1 |
| Nov 5, 2026 | 1,108,844,042,845 | 1 |
| Nov 4, 2026 | 1,018,531,639,718 | 1 |
| Nov 3, 2026 | 1,019,617,389,126 | 1 |
| Nov 2, 2026 | 1,011,767,065,339 | 1 |
| Nov 1, 2026 | 747,826,323,105 | 1 |
| Oct 31, 2026 | 799,697,698,812 | 1 |
| Oct 30, 2026 | 959,885,814,181 | 1 |
| Oct 29, 2026 | 1,177,246,450,717 | 1 |
| Oct 28, 2026 | 960,227,433,475 | 1 |
| Oct 27, 2026 | 801,693,051,154 | 1 |
| Oct 26, 2026 | 800,963,159,992 | 1 |
| Oct 25, 2026 | 800,503,370,026 | 1 |
| Oct 24, 2026 | 799,960,228,162 | 1 |
| Oct 23, 2026 | 584,527,912,314 | 1 |
| Oct 22, 2026 | 640,127,467,179 | 1 |
| Oct 21, 2026 | 234,728,339,316 | 1 |
| Oct 20, 2026 | 566,291,253,816 | 1 |
| Oct 16, 2026 | 179,047,849,184 | 1 |
| Oct 15, 2026 | 1,325,113,004,745 | 1 |
| Oct 14, 2026 | 3,737,466,999,284 | 1 |
| Oct 13, 2026 | 2,966,448,461,113 | 1 |
| Oct 12, 2026 | 3,249,049,356,669 | 1 |
| Oct 11, 2026 | 2,466,863,011,089 | 1 |
| Oct 10, 2026 | 3,149,891,365,259 | 1 |
| Oct 9, 2026 | 2,587,364,754,135 | 1 |
| Oct 8, 2026 | 632,191,853,921 | 1 |
Source: OpenRouter (openrouter.ai/rankings), as of Nov 14, 2026. CC BY 4.0. Totals include only individually published routes. Tokenizers differ by provider; the aggregated “other” category is never assigned to a model.
Technical details & API
Explore specifications, supported parameters, reasoning controls, and API identifiers for each OpenRouter route.
stepfun/step-5-previewObserved - Context
- 1,000,000
- Maximum output
- 64,000
- Tokenizer
- Other
- Input
- text, image, video
- Output
- text
- Moderation
- No
- Added to OpenRouter
- 2026-10-08
https://openrouter.ai/api/v1stepfun/step-5-previewstepfun/step-5-preview-20261008Reasoning
- Required
- Yes
- Enabled by default
- Not reported
- Default effort
- medium
- Supported efforts
- high, medium, low
Supported parameters
frequency_penaltyinclude_reasoninglogprobsmax_tokensreasoningreasoning_effortresponse_formatstopstructured_outputstemperaturetoolstop_logprobstop_p
Frequently asked questions
Answers based on the model metadata and serving offers currently in the catalog.
How much does Step 5 Preview cost?
Listed rates start at $1.00 per million input tokens and $2.70 per million output tokens. Prices vary by provider and configuration; see Pricing for details.
Which providers offer Step 5 Preview?
There are 23 cataloged offers across 21 providers. Compare model IDs, prices and limits in Providers.
What is the context window?
The cataloged model context is 1,000,000 tokens. Each serving endpoint may apply a different limit.
Does it support tools and structured output?
Tool calling: Not reported. Structured output: Not reported. Support can vary by endpoint.
Resources
Catalog history · 3 recorded revisions
Metadata observations since this model was first cataloged. These are not model release versions.
- ·
1536517be82c - ·
d535b8892b3f - ·
8cbc911833bb