{"slug": "step-5-preview-a-1m-context-moe-from-stepfun-shows-up-on-openrouter", "title": "Step 5 Preview, a 1M-context MoE from StepFun, shows up on OpenRouter", "summary": "StepFun's flagship agentic model, Step 5 Preview, has appeared on OpenRouter with a 1,000,000-token context window and a sparse Mixture-of-Experts architecture of 27B active and 600B total parameters. OpenRouter lists pricing at $1.00 per million input tokens, $2.70 per million output tokens, and $0.05 per million cache-read tokens, with 99.94% uptime and 99.43% availability over the past three days. The model supports tool calling and structured outputs via a JSON schema in response_format, and is designed for extended tasks spanning large codebases and documents, with particular strength in finance.", "body_md": "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.\n\nIt is designed for extended tasks that span large codebases and documents, using tools and refining results over multiple steps.\n\nThis model is hosted by one provider. OpenRouter forwards every request to it directly — no routing decisions to make.\n\nPricing\n\nThe average price customers actually pay for this model, next to the prices providers post. Caching and discounts mean the price actually paid is often well below the listed one.\n\nPerformance\n\nThroughput is how fast the model writes (tokens per second — higher is better). Latency is total round-trip time (lower is better). TTFT is time-to-first-token — how long before you see anything appear (lower is better).\n\nUptime\n\nUptime is the percentage of the past 3 days that at least one provider was responding to requests. Availability is the percentage of time that inference was successfully served. OpenRouter continuously monitors and uses the next-best provider when one returns an error.\n\nApps\n\nPublic apps that send the most traffic to this model. Good signal for what real production workloads look like — and a hint at which use cases this model is best suited for.\n\nActivity\n\nToken volume and request traffic to this model over time.\n\nAPI\n\nDrop-in code to call this model. OpenRouter's API is OpenAI-compatible — most SDKs work by just swapping the base URL. The only thing that changes between models is the model slug below.\n\nUptime (3d)The model was reachable. Request routed to a provider.\n\n99.94%\n\nAvailability (3d)The model returned inference from any provider. Errors and empty responses count against it.\n\n99.43%\n\nAvailability over the last 3 days\n\nLast 72 hours\n\nAvailability 99.43%\n\n3 Days Ago2 Days AgoYesterdayNow\n\nAvailability over the last 24 hours\n\nOpenRouter Availability\n\n99.75%\n\nWhen an error occurs in an upstream provider, we can recover by routing to another healthy provider, if your request filters allow it. You can access per-provider uptime data programmatically through the Endpoints API. Learn more about our load balancing and customization options.\n\nOpenClaw is an open-source AI agent that connects to your messaging apps and takes real actions on your behalf, from running commands and browsing the web to managing files and sending emails.\n\nHermes Agent is an open-source, self-improving AI agent by Nous Research that runs persistently with memory across sessions, and builds reusable skills from experience. It comes with 40+ built-in tools, including web search, browser automation, and vision, plus scheduled automations and subagents.\n\nClaude Code is Anthropic's agentic coding tool that reads your entire codebase, plans and executes changes across files, runs tests, and iterates on failures, all from natural language prompts.\n\n192Mtokens\n\nFrequently asked questions\n\nWhat is Step 5 Preview?\n\nStep 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.\n\nHow much does Step 5 Preview cost?\n\nStep 5 Preview costs $1.00/M input tokens and $2.70/M output tokens, with separate rates for Cache Read at $0.05/M tokens.\n\nWhat is the context length of Step 5 Preview?\n\nStep 5 Preview has a 1,000,000 token context window. It supports up to 64,000 completion tokens.\n\nDoes Step 5 Preview support tool calling and structured outputs?\n\nYes. Step 5 Preview accepts tools for function calling. It also supports structured outputs via a JSON schema in response_format.\n\nWhat inputs and outputs does Step 5 Preview support?\n\nStep 5 Preview accepts text, images, and video as input and returns text.", "url": "https://wpnews.pro/news/step-5-preview-a-1m-context-moe-from-stepfun-shows-up-on-openrouter", "canonical_source": "https://openrouter.ai/stepfun/step-5-preview", "published_at": "2026-10-08 16:20:20+00:00", "updated_at": "2026-10-08 16:48:24.638582+00:00", "lang": "en", "topics": ["large-language-models", "ai-agents", "ai-products", "ai-infrastructure"], "entities": ["StepFun", "Step 5 Preview", "OpenRouter", "OpenClaw", "Hermes Agent", "Nous Research", "Claude Code", "Anthropic"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/step-5-preview-a-1m-context-moe-from-stepfun-shows-up-on-openrouter", "markdown": "https://wpnews.pro/news/step-5-preview-a-1m-context-moe-from-stepfun-shows-up-on-openrouter.md", "text": "https://wpnews.pro/news/step-5-preview-a-1m-context-moe-from-stepfun-shows-up-on-openrouter.txt", "jsonld": "https://wpnews.pro/news/step-5-preview-a-1m-context-moe-from-stepfun-shows-up-on-openrouter.jsonld"}}