Together AI alternatives for open-weight inference
Together AI runs open-source models on demand, with dedicated inference and GPU clusters beside it. If you mainly need per-token inference for open-weight models, the alternatives below are Fireworks AI, DeepInfra, Groq, OpenRouter and HIPS, which is in early access by invitation. Each fact about another company links to its own page.
Published by HIPS. Not affiliated with Together AI.
- Published by
- HIPS
- Sources
- 9 pages, each linked
- Checked on
- Nov 13, 2026
- HIPS access
- Early access, by invitation
We are inviting early-access interest while we develop the platform.
What Together AI does well
From Together AI's own pages.
On-demand inference for open-source models, with dedicated model inference and GPU clusters beside it.
Together AI's home page, checkedgpt-oss-120B at $0.15 input and $0.60 output per million tokens, on demand.
Together AI's pricing page, checkedSharing your data to train other models is opt-in and off by default.
Together AI's privacy and security docs, checked
Why teams look elsewhere
From Together AI's own documented terms.
By default it stores prompts and responses and may use them for product improvements; an administrator turns storage off for zero data retention.
Together AI's privacy and security docs, checkedBatch at up to 50% lower cost on selected models; other models run at standard rates, and the completion window is 24 hours.
Together AI's batch docs, checked
What to compare
HIPS's side in its own public words; Together AI's from its own pages.
| What to check | HIPS | Together AI |
|---|---|---|
| Models | Open-weight models only, the ones its public catalog lists. On HIPS's site | On-demand inference for open-source models, with dedicated model inference and GPU clusters beside it.Together AI's home page, checked |
| Batch | Batch costs 50% less than on-demand for the same model, with one price for every window from 24 hours to 7 days. On HIPS's site | Batch at up to 50% lower cost on selected models; other models run at standard rates, and the completion window is 24 hours.Together AI's batch docs, checked |
| Training on your requests | On Free/PAYG, request inputs and generated outputs may be used for training. On the lower paid plan and the higher paid plan, request-data training is off by default. On HIPS's site | Sharing your data to train other models is opt-in and off by default.Together AI's privacy and security docs, checked |
| Request data kept | Turning training off is not, by itself, a promise of zero retention. On HIPS's site | By default it stores prompts and responses and may use them for product improvements; an administrator turns storage off for zero data retention.Together AI's privacy and security docs, checked |
The alternatives
Listed in no order of merit. Each fact links to the company's own page.
Best for: Teams that want zero data retention by default, with batch and cached-input discounts.
Zero data retention by default.
Fireworks AI's pricing page, checkedBatch at 50% off input and output, for jobs that need no immediate answer.
Fireworks AI's pricing page, checkedCached input at half price or less, with no write fee.
Fireworks AI's pricing page, checked
Best for: Teams that want per-token rates for popular open models, such as the Qwen3 family.
Qwen3-32B at $0.08 input and $0.28 output per million tokens.
DeepInfra's pricing page, checkedQwen3-14B at $0.12 input and $0.24 output per million tokens.
DeepInfra's pricing page, checked
Best for: Teams for whom response speed decides, on the models Groq lists.
GPT OSS 120B at $0.15 input and $0.60 output per million tokens, at a listed 500 tokens per second.
Groq's models page, checkedBatch at a 50% discount, with a window you set from 24 hours to 7 days.
Groq's batch docs, checked
Best for: Teams that want one API across many models and providers.
One API for 600+ models from 80+ providers on its paid plans.
OpenRouter's pricing page, checkedA fee on each credit purchase: 5.5% on the Standard pay-as-you-go plan, 8% on Business.
OpenRouter's pricing page, checked
HIPS
Early access, by invitationHIPS is building a centrally managed AI inference marketplace on a distributed GPU network. It connects people buying inference with independently supplied GPU time.
Best for: Teams that want open-weight models at one published rate per model and pool, and can join through early access.
Open-weight models only, the ones its public catalog lists.
ModelsOne published rate per model and pool, with every price change dated.
PricingBatch costs 50% less than on-demand for the same model, with one price for every window from 24 hours to 7 days.
PricingOn Free/PAYG, request inputs and generated outputs may be used for training. On the lower paid plan and the higher paid plan, request-data training is off by default.
Security and Data
Rates for the same model
HIPS's rates are read from its public catalog; Together AI's from its own page, on the date shown. Each pool states its precision.
| Offer | Precision | Input | Cached input | Output | Rate as of |
|---|---|---|---|---|---|
| HIPS, Community pool | mxfp4 (fp4) | $0.0294 | Billed as input | $0.147 | Pricing, Oct 29, 2026 |
| Together AI | Not stated on its page | $0.15 | Not stated on its page | $0.60 | Together AI's pricing page, checked |
The Certified pool is not shown: HIPS's catalog does not yet state the precision it serves this model at.
Where Together AI fits better
- You want training on your data off unless you opt in, on every plan. On HIPS's Free/PAYG, request inputs and outputs may be used for training.
- You need a model HIPS's catalog does not list.
- You want dedicated inference beside on-demand use, from one provider, today.
- You want to sign up and send a request today. HIPS is in early access, by invitation.
Moving to HIPS
Keep the client patterns you know. For supported OpenAI-style integrations, the intended starting point is three settings: your HIPS API key, base URL, and model.
Our migration guide will help you choose a model, check the features your application uses, and compare results before moving traffic.
A shared client format does not mean every provider feature or model behavior is identical.
Questions about Together AI alternatives
Does HIPS publish a rate for each model?
Yes. The Pricing page lists each language model's input, cached input and output rate in each pool, read from HIPS's public catalog, and dates every price change.
What is the difference between HIPS's two pools?
The Community pool runs on GPUs contributed by anyone, homes and companies alike, checked by HIPS before paid work. The Certified pool runs on machines HIPS has vetted and contracted to handle customer data and customer models.
Are my requests used for training on HIPS?
On Free/PAYG, request inputs and generated outputs may be used for training. On the lower paid plan and the higher paid plan, request-data training is off by default. Customer-owned model weights will never be used to train other models on any plan.
Help shape what comes next
Tell us you’re interested in HIPS. We’ll send updates about early-access availability and the capabilities we’re preparing.