Providers /Decart
Models
3
Regions
—
00
Editorial by LLMap · updated Jul 28, 2026Our take
A niche inference host with a tiny, tightly-curated catalogue of just 3 tracked offers, focused on two model families with measured throughput but sparse operational transparency.
Use this when you need one of these two specific models and throughput is a known requirement. Pick it for evaluating throughput-sensitive deployments where 52–57 TPS is relevant. Skip it if you need broad model choice, verified compliance credentials, or a price advantage on the cheaper-input option.
Strengths
- Published throughput numbers for all tracked offers — 57 TPS on z-ai-glm-5-2 and 52 TPS on moonshotai-kimi-k2-6.
- Throughput data is available for every offer in the catalogue, which is uncommon among hosts we track.
Trade-offs
- Extremely limited model selection: only 3 tracked offers, effectively 2 unique models.
- Operational transparency gaps: HQ country, SOC 2, EU endpoint, zero-retention option, and training policy are all unverified in our data.
- No price advantage on the cheaper-input model: z-ai-glm-5-2 at $1.20/$2.50 is more expensive on both input and output than moonshotai-kimi-k2-6 at $0.66/$3.40, with no benchmark differentiation.
01
Privacy & data handling
- ✓
- Yes
- ✗
- No
- No answer on record
The colour says whether the answer favours you, not whether it is a yes — not training on your prompts earns a green cross, no zero-retention option earns a red one.
Trains on your prompts
✗No
Logs prompts
✗No
✓Offered
No answer on record
Prompt retention: none.
02