vs
Together AI vs StepFun
StepFun is a Shanghai lab with cheap multimodal models, many under Apache 2.0. Together is a host with a broad catalog, training and production serving.
By The Subconscious Team · Updated
Together AI vs StepFun: key differences
StepFun builds its own models, and its workhorse Step 3.7 Flash is a 198B mixture-of-experts vision-language model with 11B active parameters, 256K context, selectable reasoning and an Apache 2.0 license. StepFun's API prices it at $0.20 in and $1.15 out, and OpenRouter carries it too. Together is not a lab. It hosts other labs' open weights, with flagships such as Kimi K3, DeepSeek V4, GLM 5.2 and Qwen 3.8, and adds serverless, batch, dedicated deployments, clusters and fine-tuning around them.
StepFun's first-party inference is hosted in China and its Western distribution and support are thin, which limits it for many buyers. It also trails frontier models on hard multimodal reasoning. Together is the stronger production platform, while StepFun is the source for low-cost vision and video understanding with small active parameters. Because Step 3.7 Flash's weights are open and run on vLLM and SGLang, teams that like the model can self-host it on Together's GPU clusters rather than calling the China-hosted API.
What Together AI and StepFun do
Together AI
Together AI is the broadest open-model platform in the category. One bill covers per-token serverless inference, batch at up to 50% off, provisioned throughput with a 99% SLA, dedicated deployments, raw GPU clusters, managed fine-tuning and code sandboxes for agents. The text catalog runs past thirty open models, including DeepSeek V4, Kimi K3, GLM 5.2, Qwen 3.8 and MiniMax M3, plus image, video, speech and embedding models. Token prices sit at parity with Fireworks and Baseten.
Example models: Kimi K3, DeepSeek V4 Pro
Full Together AI 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 Together AI or StepFun?
Together AI
Choose Together AI for
- Production serving with SLAs and rollout controls
- Frontier-scale open models for hard reasoning
- Reserved GPUs to self-host open weights like Step 3.7 Flash
StepFun
Choose StepFun for
- Cheap image and video understanding in agents
- Apache 2.0 weights with only 11B active parameters
- First-party access to new Step releases
Together AI vs StepFun at a glance
| Attribute | Together AI | StepFun |
|---|---|---|
| Model access | Open weights | Open (Apache 2.0) and API models |
| Flagship models | Kimi K3, DeepSeek V4, GLM 5.2, Qwen 3.8 | Step 3.7 Flash, Step3 |
| Speed | 0.99s TTFT on DeepSeek V4 Pro | ~128 tok/s on Step 3.7 Flash |
| Price | Parity with Fireworks and Baseten | $0.20 in, $1.15 out (Step 3.7 Flash) |
| Customization | LoRA and full SFT; RL in beta | Open weights to fine-tune |
| Deployment | Serverless, dedicated, GPU clusters | First-party API, OpenRouter |
| Long context | 512K on DeepSeek V4 Pro | 256K |
Frequently asked questions
What is the difference between Together AI and StepFun?
StepFun is a Shanghai lab with cheap multimodal models, many under Apache 2.0. Together is a host with a broad catalog, training and production serving.
When should I choose Together AI over StepFun?
Production serving with SLAs and rollout controls; Frontier-scale open models for hard reasoning; Reserved GPUs to self-host open weights like Step 3.7 Flash.
When should I choose StepFun over Together AI?
Cheap image and video understanding in agents; Apache 2.0 weights with only 11B active parameters; First-party access to new Step releases.
Is Together AI or StepFun cheaper?
Together AI: Parity with Fireworks and Baseten. StepFun: $0.20 in, $1.15 out (Step 3.7 Flash). The cheaper choice depends on the model and workload.
Which has more context, Together AI or StepFun?
Together AI: 512K on DeepSeek V4 Pro. 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.