Our take
Ionstream is a small, budget-oriented host for downloadable AI models with just five tracked offers. It often undercuts larger providers on price, but offers thin operational transparency and unverified compliance credentials.
Use this for price shopping on specific models where Ionstream beats alternatives — verify live pricing before committing. Pick it for qwen-qwen3-coder-next if 58 tokens per second meets your latency needs and the $0.11/$0.80 per million tokens price point is competitive. Skip it if you need a broad model catalogue, verified compliance, or guaranteed data residency.
- Competitive pricing on select models — deepseek-deepseek-v4-flash at $0.14/$0.28 per million tokens, qwen-qwen3-coder-next at $0.11/$0.80.
- One model with measured throughput: qwen-qwen3-coder-next at 58 tokens per second.
- Very small catalogue limits model choice — only 5 tracked offers.
- Most offers lack throughput data: 3 of 5 models have no measurement.
- Operational transparency is minimal — HQ country, EU endpoint, SOC 2, zero retention, and training policy all unverified.
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.
Prompt retention: none.