Skip to main content
AI Choice Engine

StepFun

Step 5 Preview

Step 5 Preview is a StepFun model with 1M context, $1 in / $2.70 out per million tokens, last verified October 25, 2026. Open weights: not verified. Public-board scores are linked to the publisher; vendor-only figures stay in the watch-out, not on the board.

Step 5 Preview is StepFun’s hosted text/image/video reasoning model for long-context agent, software-engineering and professional knowledge work.

Recorded profile verification October 25, 2026Individual provider fields may changeAccess: preview — Current official API guide confirms the Preview ID and capabilities. The September 18 date comes from an earlier Artificial Analysis listing, not a dated official launch notice. Current downloadable weights, licence and exact parameter counts are unverified; a past open-weight promise does not establish delivery.

AI model specification details
SpecificationStep 5 Preview
ProviderStepFun
TierFrontier
Context window1M
Max outputNot verifiedUnverified
Input / 1M tokens$1
Output / 1M tokens$2.70
WeightsNot verifiedUnverified
ParametersNot verifiedUnverified
Reasoning levelslow, medium, high
Modalitiestext, image, video
API model idstep-5-preview
ReleasedSeptember 18, 2026

Pricing tiers: StepFun hosted USD/MTok input/output $1.00/$2.70, cached input $0.10 (10% of cache-miss input), confirmed by both pricing and prompt-cache guides. First writes use cache-miss input pricing; output includes reasoning and final-answer tokens. Caching uses 256-token units and LRU eviction, without guaranteed hits. Dedicated guide advertises 1M context/input and 64K output; exact integer output ceiling is unverified. max_tokens=INF is an omitted-setting behavior, not permission to exceed context/output limits. Search/image-search/storage have separate charges.

Publisher evidence

Published benchmark results

Each result keeps its source date visible. Dates refer to source publication or capture unless the source explicitly states when it measured the result. A missing benchmark is not treated as a zero.

Published benchmark results for Step 5 Preview
BenchmarkScoreSource dateSource
Artificial Analysis Intelligence IndexStep 5 Preview442026-10-01Artificial Analysis · View source

Best for

  • Hosted multimodal agent workflows
  • Long-context API evaluation
  • Budgeted reasoning experiments

Watch out

Preview: current downloadable weights/licence and exact parameter counts remain unverified; the script-only marketing page is not proof of the older 600B/27B or October 15 release claims. Low/medium/high effort uses reasoning_effort (Chat) or output_config.effort (Messages); medium is recommended, not an established omitted default. Image/video transport limits and the shared context budget apply. Up-to-60-image, <128MB video URL and <5-minute recommendation are request conditions, not benchmark configurations. Vendor launch scores are not independent matched leaderboard rows.

History

Step 5 Preview timeline

Release and status events for this model that passed two-source verification — dated, sourced, and linked.

  • Step 5 PreviewLimited previewSeptember 18, 2026

    StepFun opens API access to Step 5 Preview ($1.00/$2.70 per MTok, 1M context) and promises open weights on 2026-10-15.

    StepFun platform pricing (step-5-preview)

Common questions

Step 5 Preview

Answered from the verified figures on this page rather than general guidance.

How much does Step 5 Preview cost per million tokens?
Step 5 Preview lists base rates of $1 per million input tokens and $2.70 per million output tokens. Estimate both input and output token volume and any additional charges for your workload. StepFun hosted USD/MTok input/output $1.00/$2.70, cached input $0.10 (10% of cache-miss input), confirmed by both pricing and prompt-cache guides. First writes use cache-miss input pricing; output includes reasoning and final-answer tokens. Caching uses 256-token units and LRU eviction, without guaranteed hits. Dedicated guide advertises 1M context/input and 64K output; exact integer output ceiling is unverified. max_tokens=INF is an omitted-setting behavior, not permission to exceed context/output limits. Search/image-search/storage have separate charges.
What is Step 5 Preview's context window?
Step 5 Preview accepts about 1M tokens of context. That only matters if you routinely send very long documents, large codebases, or multi-turn histories that approach that limit.
What is Step 5 Preview best for?
Step 5 Preview is a frontier tier from StepFun. It suits hosted multimodal agent workflows, long-context api evaluation, budgeted reasoning experiments. Preview: current downloadable weights/licence and exact parameter counts remain unverified; the script-only marketing page is not proof of the older 600B/27B or October 15 release claims. Low/medium/high effort uses reasoning_effort (Chat) or output_config.effort (Messages); medium is recommended, not an established omitted default. Image/video transport limits and the shared context budget apply. Up-to-60-image, <128MB video URL and <5-minute recommendation are request conditions, not benchmark configurations. Vendor launch scores are not independent matched leaderboard rows.