HeyGen
How a leading AI video platform unlocked new product capabilities by replacing expensive managed providers with a single, cost-effective inference API.
Together AI alternative
The Together AI alternative: open LLMs for up to 77% less.
The same open models billed per token, Serverless GPUs billed per second, and image, video and audio on the same API.
77% lowerOn gpt-oss-120b
45% lowerOn DeepSeek-V4-Flash-0731
30% lowerB300 Serverless vs on-demand
Together AI pricing vs Runware
The same open models, per 1M tokens. The biggest gap: gpt-oss-120b is 79% lower on input and 77% on output.
Price view
Together AI prices from together.ai/pricing (list prices for serverless inference, Dedicated Inference and GPU Clusters; no promotional or preemptible rates), 11 Oct, 2026. Runware prices from runware.ai model pages and Serverless pricing, 23 Oct, 2026.
LLMs per 1M tokens; images per image at about 1 MP; GPUs per GPU-hour. Together reserved rates are its 91–180 days term; Runware reserved rates require a 1–24 month term.
Build your switch estimate
Add the models you run, set monthly tokens or images, and see the same month on Runware and on Together AI.
Models priced lower on Runware
Add the ones you run, set a monthly volume, and see the two bills side by side.
ModelRunwareTogether AIChat mix
Language models · USD per 1M tokens, input / output
- gpt-oss-120bopenai:gpt-oss@120b$0.032 / $0.14$0.15 / $0.60−78%
- Gemma 4 31Bgoogle:gemma@4-31b$0.102 / $0.297$0.39 / $0.97−72%
- DeepSeek-V4.1-Flashdeepseek:v4.1@flash$0.15 / $0.60$0.30 / $1.20−50%
- DeepSeek-V4-Flash-0731deepseek:v4@flash$0.076 / $0.153$0.14 / $0.28−46%
- GLM-5.2$0.80 / $2.55$1.40 / $4.40−42%
- GLM-5.3$1.20 / $4.00$1.40 / $4.40−12%
- GLM-5.3-Flash$0.15 / $0.50$0.15 / $0.50Same price
- MiniMax M3minimax:m3@0$0.30 / $1.20$0.30 / $1.20Same price
- MiniMax M2.7minimax:m2.7@0$0.30 / $1.20$0.30 / $1.20Same price
- Kimi K3moonshotai:kimi@k3$3.00 / $15.00$3.00 / $15.00Same price
Image models · USD per image
- FLUX.2 [dev]1024×1024 · 20 steps$0.009$0.0154−42%
- Qwen-Image-2.0Per image$0.035$0.04−12%
- Qwen-Image-2.0-ProPer image$0.075$0.08−6%
- FLUX.2 [pro]1 MP$0.03$0.03Same price
- FLUX.2 [max]1 MP$0.07$0.07Same price
- FLUX.1.1 [pro]Per image$0.04$0.04Same price
- FLUX.1 Kontext [pro]Per image$0.04$0.04Same price
- FLUX.1 Kontext [max]Per image$0.08$0.08Same price
- Ideogram 3.0Default quality$0.06$0.06Same price
- Ideogram 4.0Default quality$0.06$0.06Same price
- Seedream 4.0Per image$0.03$0.03Same price
- Wan2.6 ImagePer image$0.03$0.03Same price
- Qwen-Image1024×1024 · 20 steps$0.0058$0.0058Same price
- Juggernaut Lightning Flux1024×1024$0.0017$0.0017Same price
Your monthly saving vs Together AI
$0/mo
Add models to build your estimate.
Every open LLM we serve
The full LLM price list, and how the API works
The tables here cover the models both platforms list. Every language model on Runware, with its per-token rate, is on the pricing page.
Together AI GPU pricing vs Runware Serverless, per GPU-hour
Your own models on Runware Serverless, billed per second. Runware is 19–33% lower per GPU-hour than Together; H100 on-demand is level with Together's cluster rate.
Together AI Dedicated Inference and GPU Clusters on-demand rates, and the 91–180 days reserved rate, from together.ai/pricing, 11 Oct, 2026.
The saving compares Runware Serverless with the lower of Together's two on-demand rates. Runware rates are our published Serverless pricing; reserved rates require a 1–24 month term, and GPU list rates are subject to change.
Worked example
One B300 for a month, around the clock
What 730 hours on a single B300 costs per month, at the list rates in the table above: Together AI's lower on-demand rate against Runware Serverless and Runware reserved. Reserve a baseline and burst above it on Serverless to land between the two Runware figures.
$7,293Together AI on-demand
$5,110Runware Serverless
$2,504Runware reserved
Image models, per image
FLUX and the other image models both of us serve, per image at about one megapixel.
Need something else?
Fine-tuning, rack-scale systems and video, and where to go for them on Runware.
Why teams switch
A Together AI alternative for open LLMs, GPUs and media
The same per-token API for open LLMs, Serverless GPUs for your own models, and 400K+ image, video, audio and 3D models behind the same key.
Migration
Moving from the Together AI API to Runware
For LLMs, three values change: the base URL, the API key and the model id. The OpenAI SDK, the request shape and the streaming format stay as they are. For a model on a dedicated endpoint, bring the same container to Serverless.
- Step 01
Point the base URL at api.runware.ai/v1
Together's OpenAI-compatible endpoint and Runware's accept the same request. Keep the OpenAI SDK and change base_url.
- Step 02
Swap the API key
Create a key in the Runware dashboard and put it where the Together key was.
- Step 03
Change the model id, or bring the container
Runware ids are AIR ids, listed on every model page: openai/gpt-oss-120b becomes openai:gpt-oss@120b. Running your own model on a dedicated endpoint? Deploy the same container to Serverless.
- Migrate to RunwareModels API introductionOpenAI compatibility referenceServerless quickstart
openai sdk · python
from openai import OpenAI
client = OpenAI(
api_key="your_runware_api_key",
base_url="https://api.runware.ai/v1",
)
response = client.chat.completions.create(
model="openai:gpt-oss@120b",
messages=[{"role": "user", "content": "Hi"}],
stream=True,
)No trainingOn your data
SSO + SAMLEnterprise auth
SOC 2Certified
ISO 27001Certified
GDPRCompliant
24/7Engineering support
Teams that switched and scaled
Lower bill, one API for every modality and your own models on Serverless. Here's what moving to Runware looked like for teams shipping at production scale.
Are running with us
FAQ
Together AI alternative, answered
Yes. Runware serves the same open LLMs, gpt-oss, DeepSeek, Gemma, GLM, MiniMax and Kimi among them, per token through an OpenAI-compatible endpoint. The same account runs your own models on Serverless GPUs billed per second, and 400K+ image, video, audio and 3D models behind the same key.
Per 1M tokens: gpt-oss-120b is 79% lower on input and 77% on output; DeepSeek-V4-Flash-0731 is 46% lower on input and 45% on output; Gemma 4 31B is 74% lower on input and 69% on output. GLM-5.3-Flash, MiniMax M3, MiniMax M2.7 and Kimi K3 are the same price on both. Together prices are list prices from together.ai/pricing on 11 Oct, 2026; Runware prices are read from our model pages.
Runware Serverless runs your own containers or Python handlers on dedicated H100, H200, B200 and B300 GPUs, billed per second while a worker is allocated, scaling from zero to production and back. Against Together's on-demand and reserved cluster rates, Runware is 19–33% lower per GPU-hour on the GPUs in the table. Reserved capacity is on a 1 to 24 month term. If you need the machines themselves, with your own Kubernetes stack or SSH access, use GPU Compute. GB200 NVL72 and GB300 NVL72 racks are scoped with our engineers: talk to us.
Not as a managed LLM fine-tuning service today. Runware Serverless runs training and fine-tuning jobs on your own code, on single or multi-GPU workers, and the Models API offers style LoRA training for image models. If you rely on Together's managed LLM fine-tuning, get in touch and tell us what you train.
If you call Together through its OpenAI-compatible endpoint, yes. Runware's /v1/chat/completions speaks the same protocol, streaming and reasoning_content included. Change the base URL, the API key and the model id; see the compatibility reference. If you use Together's own SDK, switch the client to the OpenAI SDK first; the request shape is the same.
The same account runs 400K+ image, video, audio and 3D models: FLUX.2, Qwen Image, Ideogram, Seedream, Nano Banana and Kontext for images, billed per image. The newest video models, Veo 3.1, Kling VIDEO 3.0 Pro, Seedance 2.5 and Wan3.0 among them, are on the same key. Browse the model directory for the model you run.
Volume pricing and reserved capacity
Running billions of tokens or a GPU fleet? Talk to us.
Everything above is our published rate. Running billions of tokens a month, or a GPU fleet? Talk to us about volume pricing and reserved capacity.