Greenference
Explore Greenference models on TrustedRouter with current routes, token pricing, policy sources, privacy posture, regional availability, and API support.
Greenferencegreenference
These privacy labels describe Greenference, the upstream model provider. Confidential requires all three: verified provider compute, end-to-end encryption and explicit zero data retention. Account and billing metadata are outside the content-retention claim.
| Provider | Greenference |
|---|---|
| API operator country | FR Legal home of the API operator, not an inference-location guarantee. |
| Headquarters | Not verified |
| Routing status | Active |
| Provider website | https://greenference.com/ |
| Models | 10 public models |
| Credits routes | 10 |
| Zero data retention | yes All routes |
| Verified confidential inference | Not verified |
| Policy note | Greenference SAS's DPA commits to EU-only inference, no persistent prompt/output storage and no training or secondary use. Shared prefix caches remain in volatile GPU memory until eviction or instance restart; they are not account-isolated and have no fixed maximum lifetime. Content-free operational metadata is retained. Greenference runs its own serving software on rented EU compute; this is contractual ZDR, not verified confidential compute or E2EE. Its current electricity mix is unverified, not a 100% renewable-energy commitment. Policy source |
Inference locations
Provider API declarations are also available on each model's endpoint API. Direct chat and Responses replies separate advertised availability from serving location in usage.inference_location; the actual serving region remains unknown when not reported.
Where the upstream GPUs execute the model, not the provider's headquarters, API ingress, or data storage. ZDR and confidential inference do not by themselves restrict geography.
- Declared countries or regions
- European Union (countries unspecified)
All inference is on Greenference-operated serving software on rented EU GPU capacity. - Can locations change?
- Dynamic within the EU; requests fail rather than leave the EU. Exact countries are not disclosed.
- Provider region pinning
- EU-only service; no country-specific pin or actual-served-region response field.
- Pinning through TrustedRouter
- Select provider.only=[greenference] with allow_fallbacks=false to restrict this route.
- Infrastructure declaration
- EU-only inference, dynamically placed; no individual-country commitment.
- Evidence
- Public DPA and provider application; authenticated Catalog v2. Reviewed .
Provider selection alone is not a region pin. TrustedRouter's US provider filter selects company jurisdiction, not GPU location. A regional gateway hostname does not pin a downstream GPU. For a binding residency requirement, confirm the model, region, failover scope and enforced controls before sending sensitive data.
Measured performance
24 samplesContinuously sampled across Greenference's routed models: p50 TTFT, effective throughput, and success rate. Effective throughput uses provider-reported output tokens over complete request time. Unsupported route and probe-configuration rows are separated from provider downtime. No prompt or output content stored.
| p50 TTFT | 1389 ms |
|---|---|
| Effective throughput | 27 tok/s n=1 |
| Uptime | 100.00% |
| Model | p50 TTFT | Effective throughput | Uptime | Config excluded | Availability samples |
|---|---|---|---|---|---|
| z-ai/glm-5.3-flash | 1170 ms | — | 100.00% | — | 3 |
| qwen/qwen3.5-9b | 1221 ms | — | 100.00% | — | 2 |
| openai/gpt-oss-20b | 1274 ms | — | 100.00% | — | 2 |
| qwen/qwen3-30b-a3b | 1367 ms | — | 100.00% | — | 1 |
| z-ai/glm-4.7-flash | 1389 ms | — | 100.00% | — | 5 |
| google/gemma-4-31b-it | 1573 ms | 27 tok/s n=1 | 100.00% | — | 1 |
| qwen/qwen3-8b | 2041 ms | — | 100.00% | — | 2 |
| qwen/qwen3.6-27b | 2187 ms | — | 100.00% | — | 5 |
| qwen/qwen3-32b | 2255 ms | — | 100.00% | — | 2 |
| qwen/qwen3-14b | 2438 ms | — | 100.00% | — | 1 |
Greenference performance history · Full provider & model leaderboard.
Models served by Greenference.
Each row links to pricing, provider, benchmark, and API pages for the model.
| Model | AI IQ | Context | Input | Cached input | Output |
|---|---|---|---|---|---|
google/gemma-4-31b-itGoogle: Gemma 4 31B |
IQ 101#110 | 262,144 | $0.156668/1M | $0.060927/1M | $0.4431/1M |
openai/gpt-oss-20bOpenAI: gpt-oss-20b |
IQ 94#140 | 131,072 | $0.01/1M | $0.01/1M | $0.047475/1M |
qwen/qwen3-14bQwen: Qwen3 14B |
— | 40,960 | $0.08018/1M | $0.04009/1M | $0.24054/1M |
qwen/qwen3-30b-a3bQwen: Qwen3 30B A3B |
— | 40,960 | $0.100225/1M | $0.050113/1M | $0.300675/1M |
qwen/qwen3-32bQwen: Qwen3 32B |
— | 40,960 | $0.100225/1M | $0.050113/1M | $0.300675/1M |
qwen/qwen3-8bQwen: Qwen3 8B |
— | 41,000 | $0.471058/1M | $0.235529/1M | $0.471058/1M |
qwen/qwen3.5-9bQwen: Qwen3.5-9B |
IQ 98#131 | 262,144 | $0.0422/1M | $0.0211/1M | $0.068575/1M |
qwen/qwen3.6-27bQwen: Qwen3.6 27B |
IQ 111#72 | 262,144 | $0.15825/1M | $0.079125/1M | $1.42425/1M |
z-ai/glm-4.7-flashZ.ai: GLM 4.7 Flash |
— | 131,072 | $0.07385/1M | $0.01512/1M | $0.429976/1M |
z-ai/glm-5.3-flashZ.ai: GLM 5.3 Flash |
IQ 120#42 | 1,048,576 | $0.1477/1M | $0.02954/1M | $0.51695/1M |
Questions
Does Greenference have zero data retention?
TrustedRouter records Greenference as supporting provider-level zero data retention based on the policy source linked on this page. This is a provider policy claim, separate from TrustedRouter's content-stateless real-time gateway and from end-to-end confidential compute.
Is Greenference end-to-end encrypted?
TrustedRouter does not currently mark Greenference as end-to-end encrypted at the provider boundary. The TrustedRouter gateway is still attested, but the selected provider normally receives the request in order to run the model. Use trustedrouter/e2e for the stronger route requirement.
Which Greenference models are available through TrustedRouter?
This page currently lists 10 public Greenference models, with live pricing, route count, context length, measured performance when available, and links to each model's provider and benchmark pages.