vs
Morph vs StepFun
Not substitutes. StepFun makes general multimodal models; Morph makes a small model that applies the code edits a general model writes.
By The Subconscious Team · Updated
Morph vs StepFun: key differences
StepFun is a model lab. Its Step 3.7 Flash is a 198B mixture-of-experts vision-language model with 11B active parameters, 256K context, tool use and structured outputs, priced at $0.20 in and $1.15 out per million under an Apache 2.0 license. Morph is a specialist. Its 7B Fast Apply model does one job, merging a frontier model's partial edit into a full file at 10,500+ tokens per second, and its WarpGrep, Compact and Reflex models handle search, compression and classification around a main model.
A cost-sensitive coding agent could pair them: Step 3.7 Flash as the cheap planning model and Morph as the apply step, so the planner only writes changed lines. Morph reports about 40% fewer tokens than full-file rewrites. StepFun's limits are that it trails frontier models on hard reasoning and its first-party inference is China-hosted, though its open weights run on vLLM and SGLang anywhere. Morph's limit is scope. It cannot be the main model, and its 2 to 4% merge error rate calls for tests.
What Morph and StepFun do
Morph
Morph builds small, very fast specialist models that sit beside a big coding model inside an agent. Its flagship is Fast Apply. The frontier model writes only the changed lines with // ... existing code ... markers, and Morph merges them into the full file at 10,500+ tokens per second with up to 98% accuracy. It is the same idea behind Cursor's instant apply, offered as an OpenAI-compatible API.
Example models: morph-v3-fast, morph-v3-large
Full Morph 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 Morph or StepFun?
Morph vs StepFun at a glance
| Attribute | Morph | StepFun |
|---|---|---|
| Model access | Specialist models | Open (Apache 2.0) and API models |
| Flagship models | morph-v3-fast, morph-v3-large | Step 3.7 Flash, Step3 |
| Speed | 10,500+ tok/s Fast Apply | ~128 tok/s on Step 3.7 Flash |
| Price | ~40% fewer tokens than full rewrites | $0.20 in, $1.15 out (Step 3.7 Flash) |
| Customization | Fine-tuning offered | Open weights to fine-tune |
| Deployment | OpenAI-compatible API | First-party API, OpenRouter |
| Long context | Unknown | 256K |
Frequently asked questions
What is the difference between Morph and StepFun?
Not substitutes. StepFun makes general multimodal models; Morph makes a small model that applies the code edits a general model writes.
When should I choose Morph over StepFun?
The apply step in any coding agent; Fast repository search with WarpGrep; Compressing context before the main model reads it.
When should I choose StepFun over Morph?
A cheap multimodal main model for agents; Self-hosting Apache 2.0 weights with 11B active; Image and video understanding in 256K context.
Is Morph or StepFun cheaper?
Morph: ~40% fewer tokens than full rewrites. StepFun: $0.20 in, $1.15 out (Step 3.7 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.