Subconscious
vs

Amazon Bedrock vs StepFun

StepFun makes cheap multimodal Step models, many Apache 2.0, served from China. Bedrock serves 100+ models inside AWS security. Low-cost open multimodal versus governed breadth.

By The Subconscious Team · Updated

Amazon Bedrock vs StepFun: key differences

StepFun is a Shanghai lab focused on efficient multimodal models. Step 3.7 Flash is a 198B mixture-of-experts vision-language model with just 11B active, 256K context and Apache 2.0 weights, priced at $0.20 in and $1.15 out on StepFun's own API. Its first-party inference is China-hosted and Western support is thin. Bedrock is AWS's managed model service, with 100+ models from 18+ providers, closed and open, under IAM, PrivateLink, KMS and CloudTrail. Its Step3 model used attention and disaggregation techniques designed to stay cheap on low-end accelerators.

The open license gives teams a route between the two. Step weights run on vLLM and SGLang, and Bedrock's Custom Model Import accepts open-weight checkpoints, which could keep a Step model inside AWS controls. Whether a given Step checkpoint is supported is worth checking. For cheap image and video understanding with no residency concerns, StepFun's API is the lowest-effort option. For regulated workloads or hard multimodal reasoning, where Step trails the frontier, Bedrock's closed models fit better.

What Amazon Bedrock and StepFun do

Amazon Bedrock

Amazon Bedrock is AWS's managed model service and has become the default AI control plane for many enterprises. One API reaches 100+ models from 18+ providers, including Anthropic's Claude family, Meta, Mistral, DeepSeek, Amazon's own Nova models, and, since an April 2026 partnership expansion, OpenAI models up to GPT-6 Astra. Switching models is usually just a new model ID. Every call inherits IAM, PrivateLink, KMS encryption and CloudTrail logging, and provider models never train on customer data.

Example models: Claude Opus, GPT-6 Astra

Full Amazon Bedrock profile

StepFun

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 profile

Should you choose Amazon Bedrock or StepFun?

Amazon Bedrock

Choose Amazon Bedrock for

  • Regulated multimodal work on frontier closed models.
  • Importing open checkpoints under AWS controls.
  • Enterprise procurement through AWS.

StepFun

Choose StepFun for

  • Cheap vision and video understanding.
  • Apache 2.0 weights with 11B active parameters.
  • 256K context on a low-cost API.

Amazon Bedrock vs StepFun at a glance

AttributeAmazon BedrockStepFun
Model accessClosed and open, 100+ modelsOpen (Apache 2.0) and API models
Flagship modelsClaude, GPT-6 Astra, Nova, DeepSeekStep 3.7 Flash, Step3
SpeedLatency-optimized option on some models~128 tok/s on Step 3.7 Flash
Price~20–35% above direct; Claude at parity$0.20 in, $1.15 out (Step 3.7 Flash)
CustomizationFine-tuning, Custom Model ImportOpen weights to fine-tune
DeploymentManaged on AWS, AgentCoreFirst-party API, OpenRouter
Long contextVaries by model256K

Frequently asked questions

What is the difference between Amazon Bedrock and StepFun?

StepFun makes cheap multimodal Step models, many Apache 2.0, served from China. Bedrock serves 100+ models inside AWS security. Low-cost open multimodal versus governed breadth.

When should I choose Amazon Bedrock over StepFun?

Regulated multimodal work on frontier closed models; Importing open checkpoints under AWS controls; Enterprise procurement through AWS.

When should I choose StepFun over Amazon Bedrock?

Cheap vision and video understanding; Apache 2.0 weights with 11B active parameters; 256K context on a low-cost API.

Is Amazon Bedrock or StepFun cheaper?

Amazon Bedrock: ~20–35% above direct; Claude at parity. StepFun: $0.20 in, $1.15 out (Step 3.7 Flash). The cheaper choice depends on the model and workload.

Which has more context, Amazon Bedrock or StepFun?

Amazon Bedrock: Varies by model. 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.

Subconscious vs Amazon BedrockSubconscious vs StepFun