vs
OpenAI vs StepFun
Step 3.7 Flash lists almost exactly at GPT-5.6 Luna's price, with Apache 2.0 weights and vision. OpenAI counters with more context, more tiers and wider reach.
By The Subconscious Team · Updated
OpenAI vs StepFun: key differences
At the low end, the price match is close. StepFun's Step 3.7 Flash costs $0.20 in and $1.15 out per million tokens, while GPT-5.6 Luna costs $0.20 in and $1.20 out. The difference is what comes with each. Step 3.7 Flash is a 198B mixture-of-experts vision-language model with 11B active parameters, 256K context, selectable reasoning levels, tool use and structured outputs, and its weights ship under Apache 2.0 for self-hosting on vLLM or SGLang. Luna carries OpenAI's 1.05M window and the Batch and caching discounts.
Above that tier the two part ways. OpenAI runs a full ladder up to GPT-6 Astra for frontier coding and computer use, while StepFun trails frontier models on hard multimodal reasoning. StepFun's first-party inference is hosted in China and its Western distribution and support are thin, though OpenRouter carries its models. OpenAI sells direct, through Azure OpenAI and through Bedrock. For cheap vision and video understanding that a team wants to self-host, StepFun is the better value. For a Western product that may need to move up to frontier models later, OpenAI is simpler.
What OpenAI and StepFun do
OpenAI
OpenAI runs the most widely adopted closed-model API. Its September 2026 lineup has GPT-6 Astra at the top for computer use, coding and long agentic runs, priced at $10 in and $50 out per million tokens. Below it sits the GPT-5.6 family: Sol for hard professional work, Terra as the balanced default, and Luna for high-volume jobs at $0.20 in and $1.20 out. All of them carry a 1.05M token context window with up to 128K output.
Example models: GPT-6 Astra, GPT-5.6 Terra
Full OpenAI profileStepFun
StepFun is a Shanghai AI lab known for efficient multimodal models, with a mix of proprietary API models and open-weight releases. Its current 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 3.7 Flash, Step3
Full StepFun profileShould you choose OpenAI or StepFun?
OpenAI
Choose OpenAI for
- Scaling from Luna up to frontier tiers on one API
- Western enterprises buying through Azure or Bedrock
- Prompts beyond 256K tokens
StepFun
Choose StepFun for
- Cheap vision and video understanding in agents
- Self-hosting an Apache 2.0 model with 11B active parameters
- Structured outputs and tool use at a Luna-level price
OpenAI vs StepFun at a glance
| Attribute | OpenAI | StepFun |
|---|---|---|
| Model access | Closed, plus open gpt-oss | Open (Apache 2.0) and API models |
| Flagship models | GPT-6 Astra, GPT-5.6 Sol, Terra, Luna | Step 3.7 Flash, Step3 |
| Speed | Fast mode: up to 2.5x at 2x price | ~128 tok/s on Step 3.7 Flash |
| Price | $0.20–$10 in, $1.20–$50 out per 1M | $0.20 in, $1.15 out (Step 3.7 Flash) |
| Customization | N/A | Open weights to fine-tune |
| Deployment | API, Azure OpenAI, Bedrock | First-party API, OpenRouter |
| Long context | 1.05M; 2x input past 272K | 256K |
Frequently asked questions
What is the difference between OpenAI and StepFun?
Step 3.7 Flash lists almost exactly at GPT-5.6 Luna's price, with Apache 2.0 weights and vision. OpenAI counters with more context, more tiers and wider reach.
When should I choose OpenAI over StepFun?
Scaling from Luna up to frontier tiers on one API; Western enterprises buying through Azure or Bedrock; Prompts beyond 256K tokens.
When should I choose StepFun over OpenAI?
Cheap vision and video understanding in agents; Self-hosting an Apache 2.0 model with 11B active parameters; Structured outputs and tool use at a Luna-level price.
Is OpenAI or StepFun cheaper?
OpenAI: $0.20–$10 in, $1.20–$50 out per 1M. StepFun: $0.20 in, $1.15 out (Step 3.7 Flash). The cheaper choice depends on the model and workload.
Which has more context, OpenAI or StepFun?
OpenAI: 1.05M; 2x input past 272K. StepFun: 256K.
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.