vs
StepFun vs Dolphin
StepFun sells its own multimodal models through a first-party API. Dolphin serves one cheap open model on a decentralized network of consumer GPUs.
By The Subconscious Team · Updated
StepFun vs Dolphin: key differences
StepFun's Step 3.7 Flash reads images and video with 256K context and tool use, at $0.20 in and $1.15 out, and ships under Apache 2.0. First-party inference is hosted in China. Dolphin's API launched in October 2026 with one model, DeepSeek V4.1 Flash, at $0.01 in and $0.07 out per million tokens, served on idle consumer GPUs across a decentralized network. It publishes no speed, context or SLA, and signup runs through a crypto wallet and Telegram.
StepFun is the more complete option for multimodal work and has published context and speed figures. Dolphin is far cheaper per token, but its API is new and its latency less predictable, so it fits background jobs more than user-facing ones.
What StepFun and Dolphin 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. 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 profile
Dolphin
Dolphin began in 2023 as Eric Hartford's series of open-source, uncensored fine-tunes, trained to drop the refusals and baked-in guidelines of their base models so that the developer sets the rules through the system prompt. Current releases include Dolphin Llama 3.1 8B and Dolphin Mistral 24B Venice Edition, built with Venice, and the project publishes nearly 100 models on Hugging Face. It also runs a chat app, a Telegram bot and a synthetic data service that generates answers to commonly refused prompts.
Example models: DeepSeek V4.1 Flash, Dolphin Mistral 24B
Full Dolphin profileShould you choose StepFun or Dolphin?
StepFun vs Dolphin at a glance
| Attribute | StepFun |  Dolphin |
|---|---|---|
| Model access | Open (Apache 2.0) and API models | Open weights, uncensored fine-tunes |
| Flagship models | Step 5 Preview, Step 3.7 Flash | DeepSeek V4.1 Flash, Dolphin Mistral 24B |
| Speed | ~128 tok/s on Step 3.7 Flash | Unknown |
| Price | $0.20 in, $1.15 out (Step 3.7 Flash) | $0.01 in, $0.07 out (DeepSeek V4.1 Flash) |
| Customization | Open weights to fine-tune | Own uncensored fine-tunes |
| Deployment | First-party API, OpenRouter | Decentralized serverless API |
| Long context | Up to 1M (Step 5 Preview); 256K on Step 3.7 Flash | Unknown |
Frequently asked questions
What is the difference between StepFun and Dolphin?
StepFun sells its own multimodal models through a first-party API. Dolphin serves one cheap open model on a decentralized network of consumer GPUs.
When should I choose StepFun over Dolphin?
Image and video understanding; Published 256K context; Apache 2.0 weights to self-host.
When should I choose Dolphin over StepFun?
Very low prices on DeepSeek V4.1 Flash; Cheap background or batch calls; Models without vendor-imposed refusals.
Is StepFun or Dolphin cheaper?
StepFun: $0.20 in, $1.15 out (Step 3.7 Flash). Dolphin: $0.01 in, $0.07 out (DeepSeek V4.1 Flash). The cheaper choice depends on the model and workload.
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.