Our take
Together AI is a US-headquartered host for downloadable models with a strong compliance posture and measured throughput on every sampled offer. Its 27 tracked offers skew toward smaller models at competitive prices, while larger models sit at a premium.
Use this for compliance-sensitive workloads that need an independent security attestation, a no-training guarantee, and a zero-retention option. Pick it when throughput predictability matters — every sampled offer has a measured figure. Skip it if you need a verified EU endpoint, a deep catalogue, or the cheapest price on mid-size and larger models.
- Strong compliance and data-privacy posture: independent security attestation, no training on prompts, and a zero-retention option available.
- All sampled offers include measured throughput — 11 to 93 tokens per second.
- Competitive pricing on small-parameter models, with some of the cheaper tracked options in that tier.
- Limited catalogue depth versus larger open-weight hosts — 27 tracked offers.
- Premium pricing on mid-size and larger models; EU-region presence is unverified in our data.
- Inconsistent throughput for identical model listings: the same Llama 3.3 70B appears twice at the same price with different measured speeds, suggesting unexplained variance.
Point your tools here
We have not recorded what a router needs for Together AI yet — no base URL and no docs link on file. That is our gap, not a sign Together AI has no API; its own documentation is the place to look until we close it.
Privacy & data handling
These answers cover requests sent to Together AI through OpenRouter, as OpenRouter records them, last read 58 min ago. For Together AI's own API we hold no answer: check its terms. SOC 2 and the data processing agreement below are hand-checked, not part of that sync.
We have not dated this attestation or linked the report yet.
Prompt retention: none.