vs
StepFun vs RunInfra
StepFun offers Apache 2.0 multimodal models at low token prices; RunInfra offers a small hosted library, $10 coding plans and an agent that builds endpoints.
By The Subconscious Team · Updated
StepFun vs RunInfra: key differences
Both providers push cheap models, but they package them differently. StepFun is a model lab. Its Step 3.7 Flash reads images and video with 256K context, tool use and structured outputs, at $0.20 in and $1.15 out, and its weights ship under Apache 2.0. RunInfra is a host with a small eight-model library, mostly mid-size models such as Nemotron 3.5 Lightning 30B and Qwen 3.8 27B, sold per token with cached discounts or on coding plans from $10 a month that plug into Claude Code, Codex and many other CLIs.
RunInfra's deployment agent is its other half. It picks a model, benchmarks it across GPUs from L4 to B200, searches quantized variants like AWQ, GPTQ and FP8, and ships an endpoint that scales to zero, and it accepts custom uploads up to 50 GB. That could in principle host an open Step model, worth checking against the upload limit. For vision and video work, StepFun's own models are the direct path. For coding on a budget, RunInfra's plans fit. StepFun's first-party hosting is in China, and RunInfra is a young company.
What StepFun and RunInfra do
StepFun
StepFun is a Shanghai AI lab known for efficient multimodal models, with a mix of proprietary API models and open-weight releases. In September 2026 it launched Step 5 Preview, its new flagship for agentic work: a 600B-total, 27B-active mixture-of-experts model with 1M context and text and image input, which Artificial Analysis scores level with Kimi K3. It costs $1.00 in and $2.70 out per million tokens, and its weights followed on Hugging Face in mid-October under the custom StepFun Model License 1.0, which allows commercial use with restrictions. Its cheaper workhorse, Step 3.7 Flash, came out in May 2026 as a 198B mixture-of-experts vision-language model with only 11B active parameters. It has 256K context, selectable reasoning levels, tool use and structured outputs, and it ships under Apache 2.0. StepFun's own API prices it at $0.20 in and $1.15 out per million tokens, and OpenRouter carries it too.
Example models: Step 5 Preview, Step 3.7 Flash
Full StepFun profileRunInfra
RunInfra pitches open models built for agents, with two ways in. Its hosted Model APIs serve an eight-model library, including Nemotron 3.5 Lightning 30B, Qwen 3.8 27B, Ornith 1.5 35B and 122B, GLM 5.3, GLM 5.3 Flash and two versions of DeepSeek V4.1 Flash, behind one key that works with both the OpenAI and Anthropic SDKs. Prices run from $0.05 in and $0.15 out on Nemotron to $1.30 in and $4.20 out on GLM 5.3, and cached input starts at $0.01. Coding plans start at $10 a month with limits that reset every five hours and every week, and they plug into Claude Code, Codex, OpenCode, Cline, Aider and dozens of other agent CLIs.
Example models: Nemotron 3.5 Lightning 30B, Qwen 3.8 27B
Full RunInfra profileShould you choose StepFun or RunInfra?
StepFun
Choose StepFun for
- Multimodal understanding across images and video
- Apache 2.0 weights for self-hosting
- Structured outputs and tool use at low token prices
RunInfra
Choose RunInfra for
- A $10 monthly coding plan for agent CLIs
- Automated quantization and GPU selection
- Scale-to-zero endpoints with cold starts under two seconds
StepFun vs RunInfra at a glance
| Attribute | StepFun | RunInfra |
|---|---|---|
| Model access | Open weights (Apache 2.0 or StepFun license) and API models | Open weights |
| Flagship models | Step 5 Preview, Step 3.7 Flash | Nemotron 3.5 Lightning 30B, Qwen 3.8 27B |
| Speed | ~128 tok/s on Step 3.7 Flash | ~541 tok/s on Nemotron 3.5 Lightning 30B (vendor); cold starts under 2s |
| Price | $0.20 in, $1.15 out (Step 3.7 Flash) | From $0.05 in, $0.15 out; coding plans from $10 a month |
| Customization | Open weights to fine-tune | Uploads up to 50 GB; auto-quantization |
| Deployment | First-party API, OpenRouter | Model APIs, agent-built endpoints |
| Long context | Up to 1M (Step 5 Preview); 256K on Step 3.7 Flash | Varies by model |
Frequently asked questions
What is the difference between StepFun and RunInfra?
StepFun offers Apache 2.0 multimodal models at low token prices; RunInfra offers a small hosted library, $10 coding plans and an agent that builds endpoints.
When should I choose StepFun over RunInfra?
Multimodal understanding across images and video; Apache 2.0 weights for self-hosting; Structured outputs and tool use at low token prices.
When should I choose RunInfra over StepFun?
A $10 monthly coding plan for agent CLIs; Automated quantization and GPU selection; Scale-to-zero endpoints with cold starts under two seconds.
Is StepFun or RunInfra cheaper?
StepFun: $0.20 in, $1.15 out (Step 3.7 Flash). RunInfra: From $0.05 in, $0.15 out; coding plans from $10 a month. The cheaper choice depends on the model and workload.
Which has more context, StepFun or RunInfra?
StepFun: Up to 1M (Step 5 Preview); 256K on Step 3.7 Flash. RunInfra: Varies by model.
Related comparisons
Running long-horizon agents?
If your agents run past 200K tokens, compare both against Subconscious. Our inference stack treats a long-horizon trace as the primary workload, so speed, cost, and accuracy hold up deep into the trace.